{"api":"knowyourml","resource":"reports","count":68,"items":[{"title":"mixup: Training on Convex Combinations","slug":"mixup-training-convex-combinations","description":"mixup trains neural networks on convex combinations of examples and labels, improving generalization and robustness to noise.","tags":["mixup","data-augmentation","regularization","training","generalization"],"category":"training","difficulty":"intermediate","year":2017,"readingTimeMinutes":5,"paperType":"AI-generated review","published":"2026-10-04","author":"Know Your ML","sources":[{"id":"arXiv:1710.09412","title":"mixup: Beyond Empirical Risk Minimization","url":"https://arxiv.org/abs/1710.09412"},{"id":"arXiv:1512.03385","title":"Deep Residual Learning for Image Recognition","url":"https://arxiv.org/abs/1512.03385"},{"id":"arXiv:1611.03530","title":"Understanding Deep Learning Requires Rethinking Generalization","url":"https://arxiv.org/abs/1611.03530"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=mixup-training-convex-combinations"},{"title":"XLNet: Generalized Autoregressive Pretraining","slug":"xlnet-generalized-autoregressive-pretraining","description":"XLNet combines permutation language modeling with Transformer-XL recurrence, outperforming BERT on SQuAD, GLUE and RACE benchmarks.","tags":["xlnet","pretraining","transformers","nlp","permutation-language-modeling"],"category":"architectures","difficulty":"intermediate","year":2019,"readingTimeMinutes":5,"paperType":"AI-generated review","published":"2026-10-04","author":"Know Your ML","sources":[{"id":"arXiv:1906.08237","title":"XLNet: Generalized Autoregressive Pretraining for Language Understanding","url":"https://arxiv.org/abs/1906.08237"},{"id":"arXiv:1810.04805","title":"BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding","url":"https://arxiv.org/abs/1810.04805"},{"id":"arXiv:1901.02860","title":"Transformer-XL: Attentive Language Models Beyond a Fixed-Length Context","url":"https://arxiv.org/abs/1901.02860"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=xlnet-generalized-autoregressive-pretraining"},{"title":"Variational Autoencoders","slug":"vae-auto-encoding-variational-bayes","description":"Variational Autoencoders learn compressed latent representations with the ELBO objective and reparameterization trick for generative modeling.","tags":["vae","variational-inference","generative","latent-variable-models","autoencoders"],"category":"generative","difficulty":"intermediate","year":2013,"readingTimeMinutes":5,"paperType":"AI-generated review","published":"2026-10-04","author":"Know Your ML","sources":[{"id":"arXiv:1312.6114","title":"Auto-Encoding Variational Bayes","url":"https://arxiv.org/abs/1312.6114"},{"id":"arXiv:1406.2661","title":"Generative Adversarial Nets","url":"https://arxiv.org/abs/1406.2661"},{"id":"arXiv:1711.00937","title":"Neural Discrete Representation Learning","url":"https://arxiv.org/abs/1711.00937"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=vae-auto-encoding-variational-bayes"},{"title":"VGG: Very Deep Convolutional Networks","slug":"vgg-very-deep-convolutional-networks","description":"VGG showed that stacking small 3x3 convolutions into 16-19 layer CNNs wins at ImageNet classification — the simple deep-learning blueprint behind modern vision.","tags":["vgg","cnn","computer-vision","image-classification","deep-learning"],"category":"vision","difficulty":"beginner","year":2014,"readingTimeMinutes":5,"paperType":"AI-generated review","published":"2026-10-04","author":"Know Your ML","sources":[{"id":"arXiv:1409.1556","title":"Very Deep Convolutional Networks for Large-Scale Image Recognition","url":"https://arxiv.org/abs/1409.1556"},{"id":"arXiv:1409.4842","title":"Going Deeper with Convolutions","url":"https://arxiv.org/abs/1409.4842"},{"id":"arXiv:1512.03385","title":"Deep Residual Learning for Image Recognition","url":"https://arxiv.org/abs/1512.03385"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=vgg-very-deep-convolutional-networks"},{"title":"SimCLR: Simple Contrastive Learning of Visual Representations","slug":"simclr-simple-contrastive-visual-learning","description":"SimCLR learns visual representations without labels via contrastive learning on augmented image pairs, matching supervised accuracy with linear probes.","tags":["simclr","contrastive-learning","self-supervised","computer-vision","representation-learning"],"category":"vision","difficulty":"intermediate","year":2020,"readingTimeMinutes":5,"paperType":"AI-generated review","published":"2026-10-04","author":"Know Your ML","sources":[{"id":"arXiv:2002.05709","title":"A Simple Framework for Contrastive Learning of Visual Representations","url":"https://arxiv.org/abs/2002.05709"},{"id":"arXiv:1911.05722","title":"Momentum Contrast for Unsupervised Visual Representation Learning","url":"https://arxiv.org/abs/1911.05722"},{"id":"arXiv:2006.07733","title":"Bootstrap Your Own Latent: A New Approach to Self-Supervised Learning","url":"https://arxiv.org/abs/2006.07733"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=simclr-simple-contrastive-visual-learning"},{"title":"PPO: Proximal Policy Optimization","slug":"ppo-proximal-policy-optimization","description":"PPO stabilizes policy gradients with a clipped objective, becoming the default RL algorithm behind robotics, games and RLHF fine-tuning.","tags":["ppo","policy-gradient","reinforcement-learning","rlhf","trust-region"],"category":"reinforcement-learning","difficulty":"intermediate","year":2017,"readingTimeMinutes":5,"paperType":"AI-generated review","published":"2026-10-04","author":"Know Your ML","sources":[{"id":"arXiv:1707.06347","title":"Proximal Policy Optimization Algorithms","url":"https://arxiv.org/abs/1707.06347"},{"id":"arXiv:1502.05477","title":"Trust Region Policy Optimization","url":"https://arxiv.org/abs/1502.05477"},{"id":"arXiv:1602.01783","title":"Asynchronous Methods for Deep Reinforcement Learning","url":"https://arxiv.org/abs/1602.01783"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=ppo-proximal-policy-optimization"},{"title":"Masked Autoencoders Are Scalable Vision Learners","slug":"mae-masked-autoencoders-scalable-vision","description":"Masked Autoencoders pretrain Vision Transformers by reconstructing 75% masked patches, scaling self-supervised vision with simple asymmetric design.","tags":["masked-autoencoders","mae","self-supervised","vision-transformers","pretraining"],"category":"vision","difficulty":"intermediate","year":2021,"readingTimeMinutes":5,"paperType":"AI-generated review","published":"2026-10-04","author":"Know Your ML","sources":[{"id":"arXiv:2111.06377","title":"Masked Autoencoders Are Scalable Vision Learners","url":"https://arxiv.org/abs/2111.06377"},{"id":"arXiv:2010.11929","title":"An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale","url":"https://arxiv.org/abs/2010.11929"},{"id":"arXiv:1810.04805","title":"BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding","url":"https://arxiv.org/abs/1810.04805"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=mae-masked-autoencoders-scalable-vision"},{"title":"Mask R-CNN: Instance Segmentation","slug":"mask-rcnn-instance-segmentation","description":"Mask R-CNN adds a mask branch and RoIAlign to Faster R-CNN for instance segmentation — pixel-level masks, boxes and classes from one network.","tags":["instance-segmentation","mask-rcnn","object-detection","computer-vision","segmentation"],"category":"vision","difficulty":"intermediate","year":2017,"readingTimeMinutes":5,"paperType":"AI-generated review","published":"2026-10-04","author":"Know Your ML","sources":[{"id":"arXiv:1703.06870","title":"Mask R-CNN","url":"https://arxiv.org/abs/1703.06870"},{"id":"arXiv:1506.01497","title":"Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks","url":"https://arxiv.org/abs/1506.01497"},{"id":"arXiv:1612.03144","title":"Feature Pyramid Networks for Object Detection","url":"https://arxiv.org/abs/1612.03144"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=mask-rcnn-instance-segmentation"},{"title":"Layer Normalization","slug":"layer-normalization-sequence-training","description":"Layer Normalization normalizes across features instead of batches, stabilizing RNN and Transformer training at any batch size.","tags":["layer-normalization","normalization","transformers","training","rnn"],"category":"training","difficulty":"beginner","year":2016,"readingTimeMinutes":5,"paperType":"AI-generated review","published":"2026-10-04","author":"Know Your ML","sources":[{"id":"arXiv:1607.06450","title":"Layer Normalization","url":"https://arxiv.org/abs/1607.06450"},{"id":"arXiv:1502.03167","title":"Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift","url":"https://arxiv.org/abs/1502.03167"},{"id":"arXiv:1706.03762","title":"Attention Is All You Need","url":"https://arxiv.org/abs/1706.03762"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=layer-normalization-sequence-training"},{"title":"Knowledge Distillation: Teacher-Student Compression","slug":"knowledge-distillation-teacher-student","description":"Knowledge distillation compresses big teacher models into small fast students using soft targets — the key to efficient edge AI.","tags":["knowledge-distillation","compression","efficiency","teacher-student","quantization"],"category":"efficiency","difficulty":"intermediate","year":2015,"readingTimeMinutes":5,"paperType":"AI-generated review","published":"2026-10-04","author":"Know Your ML","sources":[{"id":"arXiv:1503.02531","title":"Distilling the Knowledge in a Neural Network","url":"https://arxiv.org/abs/1503.02531"},{"id":"arXiv:1910.01108","title":"DistilBERT, a distilled version of BERT","url":"https://arxiv.org/abs/1910.01108"},{"id":"arXiv:1512.03385","title":"Deep Residual Learning for Image Recognition","url":"https://arxiv.org/abs/1512.03385"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=knowledge-distillation-teacher-student"},{"title":"He Initialization: Delving Deep into Rectifiers","slug":"he-initialization-delving-into-rectifiers","description":"He initialization scales starting weights for ReLU networks, keeping signal variance stable so very deep rectifier models can train.","tags":["initialization","relu","training","deep-learning","variance-scaling"],"category":"training","difficulty":"intermediate","year":2015,"readingTimeMinutes":5,"paperType":"AI-generated review","published":"2026-10-04","author":"Know Your ML","sources":[{"id":"arXiv:1502.01852","title":"Delving Deep into Rectifiers: Surpassing Human-Level Performance on ImageNet Classification","url":"https://arxiv.org/abs/1502.01852"},{"id":"arXiv:1512.03385","title":"Deep Residual Learning for Image Recognition","url":"https://arxiv.org/abs/1512.03385"},{"id":"arXiv:1502.03167","title":"Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift","url":"https://arxiv.org/abs/1502.03167"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=he-initialization-delving-into-rectifiers"},{"title":"Grad-CAM: Visual Explanations from Deep Networks","slug":"grad-cam-visual-explanations-networks","description":"Grad-CAM explains CNN decisions with gradient-weighted heatmaps, revealing which image regions drove each prediction without retraining.","tags":["grad-cam","interpretability","explainability","computer-vision","visualization"],"category":"interpretability","difficulty":"beginner","year":2016,"readingTimeMinutes":5,"paperType":"AI-generated review","published":"2026-10-04","author":"Know Your ML","sources":[{"id":"arXiv:1610.02391","title":"Grad-CAM: Visual Explanations from Deep Networks via Gradient-based Localization","url":"https://arxiv.org/abs/1610.02391"},{"id":"arXiv:1311.2901","title":"Visualizing and Understanding Convolutional Networks","url":"https://arxiv.org/abs/1311.2901"},{"id":"arXiv:1512.04150","title":"Learning Deep Features for Discriminative Localization","url":"https://arxiv.org/abs/1512.04150"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=grad-cam-visual-explanations-networks"},{"title":"GoogLeNet and Inception: Going Deeper with Convolutions","slug":"googlenet-inception-multiscale-convolutions","description":"GoogLeNet won ILSVRC 2014 with Inception modules running multi-scale convolutions in parallel — deeper CNNs with far fewer parameters than rivals.","tags":["googlenet","inception","cnn","computer-vision","image-classification"],"category":"vision","difficulty":"beginner","year":2014,"readingTimeMinutes":5,"paperType":"AI-generated review","published":"2026-10-04","author":"Know Your ML","sources":[{"id":"arXiv:1409.4842","title":"Going Deeper with Convolutions","url":"https://arxiv.org/abs/1409.4842"},{"id":"arXiv:1409.1556","title":"Very Deep Convolutional Networks for Large-Scale Image Recognition","url":"https://arxiv.org/abs/1409.1556"},{"id":"arXiv:1512.03385","title":"Deep Residual Learning for Image Recognition","url":"https://arxiv.org/abs/1512.03385"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=googlenet-inception-multiscale-convolutions"},{"title":"Faster R-CNN: Real-Time Detection with Region Proposals","slug":"faster-rcnn-region-proposal-networks","description":"Faster R-CNN made object detection near real-time with Region Proposal Networks sharing CNN features — anchors, RoI pooling and joint training explained.","tags":["object-detection","faster-rcnn","computer-vision","region-proposals","cnn"],"category":"vision","difficulty":"intermediate","year":2015,"readingTimeMinutes":5,"paperType":"AI-generated review","published":"2026-10-04","author":"Know Your ML","sources":[{"id":"arXiv:1506.01497","title":"Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks","url":"https://arxiv.org/abs/1506.01497"},{"id":"arXiv:1504.08083","title":"Fast R-CNN","url":"https://arxiv.org/abs/1504.08083"},{"id":"arXiv:1311.2524","title":"Rich Feature Hierarchies for Accurate Object Detection and Semantic Segmentation","url":"https://arxiv.org/abs/1311.2524"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=faster-rcnn-region-proposal-networks"},{"title":"DETR: End-to-End Object Detection with Transformers","slug":"detr-end-to-end-object-detection","description":"DETR replaces anchors and NMS with a transformer predicting objects as a set — end-to-end object detection with bipartite matching, explained simply.","tags":["object-detection","detr","transformers","computer-vision","end-to-end"],"category":"vision","difficulty":"intermediate","year":2020,"readingTimeMinutes":5,"paperType":"AI-generated review","published":"2026-10-04","author":"Know Your ML","sources":[{"id":"arXiv:2005.12872","title":"End-to-End Object Detection with Transformers","url":"https://arxiv.org/abs/2005.12872"},{"id":"arXiv:1706.03762","title":"Attention Is All You Need","url":"https://arxiv.org/abs/1706.03762"},{"id":"arXiv:1506.01497","title":"Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks","url":"https://arxiv.org/abs/1506.01497"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=detr-end-to-end-object-detection"},{"title":"DDIM: Fast Deterministic Diffusion Sampling","slug":"ddim-denoising-diffusion-implicit-models","description":"DDIM turns diffusion sampling deterministic and 10-50x faster with strided steps, keeping DDPM quality while enabling latent interpolation and editing.","tags":["diffusion","ddim","sampling","generative","image-generation"],"category":"generative","difficulty":"advanced","year":2020,"readingTimeMinutes":5,"paperType":"AI-generated review","published":"2026-10-04","author":"Know Your ML","sources":[{"id":"arXiv:2010.02502","title":"Denoising Diffusion Implicit Models","url":"https://arxiv.org/abs/2010.02502"},{"id":"arXiv:2006.11239","title":"Denoising Diffusion Probabilistic Models","url":"https://arxiv.org/abs/2006.11239"},{"id":"arXiv:2105.05233","title":"Diffusion Models Beat GANs on Image Synthesis","url":"https://arxiv.org/abs/2105.05233"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=ddim-denoising-diffusion-implicit-models"},{"title":"ControlNet: Conditional Control for Diffusion Models","slug":"controlnet-conditional-control-diffusion","description":"ControlNet adds spatial conditioning like edges and pose to frozen diffusion models via zero convolutions, enabling precise controllable image generation.","tags":["controlnet","diffusion","conditioning","image-generation","generative"],"category":"generative","difficulty":"intermediate","year":2023,"readingTimeMinutes":5,"paperType":"AI-generated review","published":"2026-10-04","author":"Know Your ML","sources":[{"id":"arXiv:2302.05543","title":"Adding Conditional Control to Text-to-Image Diffusion Models","url":"https://arxiv.org/abs/2302.05543"},{"id":"arXiv:2112.10741","title":"High-Resolution Image Synthesis with Latent Diffusion Models","url":"https://arxiv.org/abs/2112.10741"},{"id":"arXiv:2302.08453","title":"T2I-Adapter: Learning Adapters to Dig Out More Controllable Ability","url":"https://arxiv.org/abs/2302.08453"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=controlnet-conditional-control-diffusion"},{"title":"Batch Normalization: Accelerating Deep Network Training","slug":"batch-normalization-accelerating-deep-training","description":"Batch Normalization normalizes layer inputs per mini-batch, enabling higher learning rates and faster, more stable deep network training.","tags":["batch-normalization","normalization","training","optimization","deep-learning"],"category":"training","difficulty":"beginner","year":2015,"readingTimeMinutes":5,"paperType":"AI-generated review","published":"2026-10-04","author":"Know Your ML","sources":[{"id":"arXiv:1502.03167","title":"Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift","url":"https://arxiv.org/abs/1502.03167"},{"id":"arXiv:1607.06450","title":"Layer Normalization","url":"https://arxiv.org/abs/1607.06450"},{"id":"arXiv:1512.03385","title":"Deep Residual Learning for Image Recognition","url":"https://arxiv.org/abs/1512.03385"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=batch-normalization-accelerating-deep-training"},{"title":"Adam: Adaptive Moment Estimation","slug":"adam-adaptive-moment-estimation","description":"Adam combines momentum with per-parameter adaptive learning rates, making it the default optimizer for training deep neural networks.","tags":["adam","optimization","training","gradient-descent","adaptive-learning-rates"],"category":"training","difficulty":"beginner","year":2014,"readingTimeMinutes":5,"paperType":"AI-generated review","published":"2026-10-04","author":"Know Your ML","sources":[{"id":"arXiv:1412.6980","title":"Adam: A Method for Stochastic Optimization","url":"https://arxiv.org/abs/1412.6980"},{"id":"arXiv:1711.05101","title":"Decoupled Weight Decay Regularization","url":"https://arxiv.org/abs/1711.05101"},{"id":"arXiv:1502.03167","title":"Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift","url":"https://arxiv.org/abs/1502.03167"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=adam-adaptive-moment-estimation"},{"title":"A3C: Asynchronous Methods for Deep Reinforcement Learning","slug":"a3c-asynchronous-methods-deep-rl","description":"A3C trains deep RL agents with parallel asynchronous workers, beating DQN on Atari in a fraction of the time on a single CPU.","tags":["a3c","actor-critic","reinforcement-learning","asynchronous-training","deep-rl"],"category":"reinforcement-learning","difficulty":"advanced","year":2016,"readingTimeMinutes":5,"paperType":"AI-generated review","published":"2026-10-04","author":"Know Your ML","sources":[{"id":"arXiv:1602.01783","title":"Asynchronous Methods for Deep Reinforcement Learning","url":"https://arxiv.org/abs/1602.01783"},{"id":"arXiv:1312.5602","title":"Playing Atari with Deep Reinforcement Learning","url":"https://arxiv.org/abs/1312.5602"},{"id":"arXiv:1707.06347","title":"Proximal Policy Optimization Algorithms","url":"https://arxiv.org/abs/1707.06347"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=a3c-asynchronous-methods-deep-rl"},{"title":"Visualizing Convolutional Networks","slug":"visualizing-cnn-zeiler-fergus","description":"Zeiler and Fergus visualized CNN layers with deconvolution, revealing edges, textures and objects emerging layer by layer inside the network.","tags":["interpretability","cnn","visualization","computer-vision","deconvnet"],"category":"interpretability","difficulty":"beginner","year":2013,"readingTimeMinutes":5,"paperType":"AI-generated review","published":"2026-09-14","author":"Know Your ML","sources":[{"id":"arXiv:1311.2901","title":"Visualizing and Understanding Convolutional Networks","url":"https://arxiv.org/abs/1311.2901"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=visualizing-cnn-zeiler-fergus"},{"title":"Tree of Thoughts","slug":"tree-of-thoughts-deliberate-problem-solving","description":"Tree of Thoughts explores multiple reasoning branches with search and self-evaluation, solving problems where single-pass chain-of-thought fails.","tags":["tree-of-thoughts","reasoning","search","planning","prompting"],"category":"applications","difficulty":"intermediate","year":2023,"readingTimeMinutes":5,"paperType":"AI-generated review","published":"2026-09-14","author":"Know Your ML","sources":[{"id":"arXiv:2305.10601","title":"Tree of Thoughts: Deliberate Problem Solving with Large Language Models","url":"https://arxiv.org/abs/2305.10601"},{"id":"arXiv:2201.11903","title":"Chain-of-Thought Prompting Elicits Reasoning in Large Language Models","url":"https://arxiv.org/abs/2201.11903"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=tree-of-thoughts-deliberate-problem-solving"},{"title":"Self-Consistency Improves Chain of Thought Reasoning in Language Models","slug":"self-consistency-decoding","description":"An AI-generated review of Self-Consistency decoding, showing how sampling multiple reasoning paths and selecting the most consistent answer improves.","tags":["reasoning","chain-of-thought","decoding","self-consistency","language-models"],"category":"applications","difficulty":"intermediate","year":2022,"readingTimeMinutes":9,"paperType":"AI-generated review","published":"2026-09-14","author":"Know Your ML","sources":[{"title":"Self-Consistency Improves Chain of Thought Reasoning in Language Models","id":"arXiv:2203.11171","url":"https://arxiv.org/abs/2203.11171"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=self-consistency-decoding"},{"title":"Segment Anything","slug":"segment-anything","description":"How the Segment Anything project reframed segmentation as a promptable task and released a 1-billion-mask dataset that made zero-shot segmentation practical.","tags":["segmentation","computer-vision","foundation-models","zero-shot","prompting"],"category":"vision","difficulty":"intermediate","year":2023,"readingTimeMinutes":6,"paperType":"AI-generated review","published":"2026-09-14","author":"Know Your ML","sources":[{"title":"Segment Anything","id":"arXiv:2304.02643","url":"https://arxiv.org/abs/2304.02643"},{"title":"An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale","id":"arXiv:2010.11929","url":"https://arxiv.org/abs/2010.11929"},{"title":"Masked Autoencoders Are Scalable Vision Learners","id":"arXiv:2111.06377","url":"https://arxiv.org/abs/2111.06377"},{"title":"Learning Transferable Visual Models From Natural Language Supervision","id":"arXiv:2103.00020","url":"https://arxiv.org/abs/2103.00020"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=segment-anything"},{"title":"Scaling Laws for Neural Language Models","slug":"scaling-laws-neural-language-models","description":"Kaplan et al. showed language model loss follows predictable power laws in parameters, data and compute — the empirical rules behind modern scaling.","tags":["scaling-laws","large-language-models","training","compute","foundations"],"category":"foundations","difficulty":"intermediate","year":2020,"readingTimeMinutes":5,"paperType":"AI-generated review","published":"2026-09-14","author":"Know Your ML","sources":[{"id":"arXiv:2001.08361","title":"Scaling Laws for Neural Language Models","url":"https://arxiv.org/abs/2001.08361"},{"id":"arXiv:2203.15556","title":"Training Compute-Optimal Large Language Models","url":"https://arxiv.org/abs/2203.15556"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=scaling-laws-neural-language-models"},{"title":"Robust Speech Recognition via Large-Scale Weak Supervision","slug":"whisper-robust-speech-recognition","description":"Whisper trained on 680,000 hours of weakly supervised web audio and showed that scale plus multitask formatting beats hand-curated speech datasets.","tags":["speech-recognition","weak-supervision","multilingual","transformers","zero-shot"],"category":"applications","difficulty":"intermediate","year":2022,"readingTimeMinutes":6,"paperType":"AI-generated review","published":"2026-09-14","author":"Know Your ML","sources":[{"title":"Robust Speech Recognition via Large-Scale Weak Supervision","id":"arXiv:2212.04356","url":"https://arxiv.org/abs/2212.04356"},{"title":"Attention Is All You Need","id":"arXiv:1706.03762","url":"https://arxiv.org/abs/1706.03762"},{"title":"wav2vec 2.0: A Framework for Self-Supervised Learning of Speech Representations","id":"arXiv:2006.11477","url":"https://arxiv.org/abs/2006.11477"},{"title":"Connectionist Temporal Classification: Labelling Unsegmented Sequence Data with Recurrent Neural Networks","id":"arXiv:1306.0543","url":"https://arxiv.org/abs/1306.0543"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=whisper-robust-speech-recognition"},{"title":"RoPE: Rotary Position Embeddings","slug":"rotary-position-embeddings","description":"Rotary position embeddings encode token order as rotations in embedding space, giving transformers relative positions that stretch to longer text.","tags":["position-embeddings","rope","transformers","long-context","architectures"],"category":"architectures","difficulty":"intermediate","year":2021,"readingTimeMinutes":5,"paperType":"AI-generated review","published":"2026-09-14","author":"Know Your ML","sources":[{"id":"arXiv:2104.09864","title":"RoFormer: Enhanced Transformer with Rotary Position Embedding","url":"https://arxiv.org/abs/2104.09864"},{"id":"arXiv:1706.03762","title":"Attention Is All You Need","url":"https://arxiv.org/abs/1706.03762"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=rotary-position-embeddings"},{"title":"ReAct: Reasoning and Acting","slug":"react-reasoning-acting-language-models","description":"ReAct interleaves chain-of-thought reasoning with tool actions like search and lookup, grounding language models in external feedback.","tags":["react","agents","tool-use","reasoning","prompting"],"category":"applications","difficulty":"beginner","year":2022,"readingTimeMinutes":5,"paperType":"AI-generated review","published":"2026-09-14","author":"Know Your ML","sources":[{"id":"arXiv:2210.03629","title":"ReAct: Synergizing Reasoning and Acting in Language Models","url":"https://arxiv.org/abs/2210.03629"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=react-reasoning-acting-language-models"},{"title":"QLoRA: Quantized Fine-Tuning","slug":"qlora-efficient-finetuning-quantized","description":"QLoRA fine-tunes 65B models on one 48GB GPU by combining 4-bit quantization with LoRA adapters — nearly matching full-precision tuning.","tags":["qlora","quantization","fine-tuning","lora","efficiency"],"category":"efficiency","difficulty":"intermediate","year":2023,"readingTimeMinutes":5,"paperType":"AI-generated review","published":"2026-09-14","author":"Know Your ML","sources":[{"id":"arXiv:2305.14314","title":"QLoRA: Efficient Finetuning of Quantized LLMs","url":"https://arxiv.org/abs/2305.14314"},{"id":"arXiv:2106.09685","title":"LoRA: Low-Rank Adaptation of Large Language Models","url":"https://arxiv.org/abs/2106.09685"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=qlora-efficient-finetuning-quantized"},{"title":"Mixtral: Sparse Mixture of Experts","slug":"mixtral-sparse-mixture-of-experts","description":"Mixtral routes each token to 2 of 8 experts, matching LLaMA-2 70B quality with 6x faster inference — open sparse MoE that actually works.","tags":["mixture-of-experts","moe","efficiency","open-models","sparse-models"],"category":"efficiency","difficulty":"intermediate","year":2024,"readingTimeMinutes":5,"paperType":"AI-generated review","published":"2026-09-14","author":"Know Your ML","sources":[{"id":"arXiv:2401.04088","title":"Mixtral of Experts","url":"https://arxiv.org/abs/2401.04088"},{"id":"arXiv:2101.03961","title":"Switch Transformers: Scaling to Trillion Parameter Models with Simple and Efficient Sparsity","url":"https://arxiv.org/abs/2101.03961"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=mixtral-sparse-mixture-of-experts"},{"title":"Mamba: Selective State Spaces","slug":"mamba-selective-state-spaces","description":"Mamba replaces attention with selective state spaces that scale linearly in sequence length — matching transformers at several times the inference speed.","tags":["mamba","state-space-models","long-context","architectures","efficiency"],"category":"architectures","difficulty":"advanced","year":2023,"readingTimeMinutes":5,"paperType":"AI-generated review","published":"2026-09-14","author":"Know Your ML","sources":[{"id":"arXiv:2312.00752","title":"Mamba: Linear-Time Sequence Modeling with Selective State Spaces","url":"https://arxiv.org/abs/2312.00752"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=mamba-selective-state-spaces"},{"title":"LLaMA: Open and Efficient Foundation Language Models","slug":"llama-open-efficient-language-models","description":"LLaMA showed a 13B model trained on a trillion tokens could beat GPT-3 175B, reframing the scaling debate around compute-optimal token budgets.","tags":["large-language-models","scaling-laws","training","open-models","efficiency"],"category":"foundations","difficulty":"intermediate","year":2023,"readingTimeMinutes":7,"paperType":"AI-generated review","published":"2026-09-14","author":"Know Your ML","sources":[{"title":"LLaMA: Open and Efficient Foundation Language Models","id":"arXiv:2302.13971","url":"https://arxiv.org/abs/2302.13971"},{"title":"Training Compute-Optimal Large Language Models","id":"arXiv:2203.15556","url":"https://arxiv.org/abs/2203.15556"},{"title":"Scaling Laws for Neural Language Models","id":"arXiv:2001.08361","url":"https://arxiv.org/abs/2001.08361"},{"title":"GLU Variants Improve Transformer","id":"arXiv:2002.05202","url":"https://arxiv.org/abs/2002.05202"},{"title":"RoFormer: Enhanced Transformer with Rotary Position Embedding","id":"arXiv:2104.09864","url":"https://arxiv.org/abs/2104.09864"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=llama-open-efficient-language-models"},{"title":"Dropout: Preventing Overfitting","slug":"dropout-preventing-overfitting","description":"Dropout randomly disables neurons during training, forcing redundant representations — a simple trick that helped make deep networks trainable.","tags":["dropout","regularization","training","overfitting","deep-learning"],"category":"training","difficulty":"beginner","year":2012,"readingTimeMinutes":5,"paperType":"AI-generated review","published":"2026-09-14","author":"Know Your ML","sources":[{"id":"arXiv:1207.0580","title":"Dropout: A Simple Way to Prevent Neural Networks from Overfitting","url":"https://arxiv.org/abs/1207.0580"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=dropout-preventing-overfitting"},{"title":"Direct Preference Optimization","slug":"direct-preference-optimization","description":"Direct Preference Optimization aligns language models from preference pairs with one stable loss — no reward model, no reinforcement learning loop.","tags":["dpo","alignment","preference-optimization","fine-tuning","rlhf"],"category":"alignment","difficulty":"intermediate","year":2023,"readingTimeMinutes":5,"paperType":"AI-generated review","published":"2026-09-14","author":"Know Your ML","sources":[{"id":"arXiv:2305.18290","title":"Direct Preference Optimization: Your Language Model is Secretly a Reward Model","url":"https://arxiv.org/abs/2305.18290"},{"id":"arXiv:1706.03741","title":"Deep Reinforcement Learning from Human Preferences","url":"https://arxiv.org/abs/1706.03741"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=direct-preference-optimization"},{"title":"Deep RL from Human Preferences","slug":"rlhf-deep-rl-from-human-preferences","description":"Christiano et al. trained agents from human preferences instead of hand-written rewards — the 2017 result that made RLHF, and ChatGPT, possible.","tags":["rlhf","alignment","human-feedback","reinforcement-learning","reward-modeling"],"category":"alignment","difficulty":"intermediate","year":2017,"readingTimeMinutes":5,"paperType":"AI-generated review","published":"2026-09-14","author":"Know Your ML","sources":[{"id":"arXiv:1706.03741","title":"Deep Reinforcement Learning from Human Preferences","url":"https://arxiv.org/abs/1706.03741"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=rlhf-deep-rl-from-human-preferences"},{"title":"Constitutional AI","slug":"constitutional-ai-harmlessness","description":"Constitutional AI trains harmless assistants from a written constitution via AI self-critique, cutting human labels while keeping helpfulness.","tags":["constitutional-ai","alignment","harmlessness","rlaif","self-critique"],"category":"alignment","difficulty":"intermediate","year":2022,"readingTimeMinutes":5,"paperType":"AI-generated review","published":"2026-09-14","author":"Know Your ML","sources":[{"id":"arXiv:2212.08073","title":"Constitutional AI: Harmlessness from AI Feedback","url":"https://arxiv.org/abs/2212.08073"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=constitutional-ai-harmlessness"},{"title":"Chinchilla: Compute-Optimal Training","slug":"chinchilla-compute-optimal-training","description":"Chinchilla proved a 70B model on 1.4T tokens beats a 280B model on the same compute — most large models were badly undertrained on data.","tags":["scaling-laws","training","chinchilla","compute-optimal","large-language-models"],"category":"foundations","difficulty":"intermediate","year":2022,"readingTimeMinutes":5,"paperType":"AI-generated review","published":"2026-09-14","author":"Know Your ML","sources":[{"id":"arXiv:2203.15556","title":"Training Compute-Optimal Large Language Models","url":"https://arxiv.org/abs/2203.15556"},{"id":"arXiv:2001.08361","title":"Scaling Laws for Neural Language Models","url":"https://arxiv.org/abs/2001.08361"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=chinchilla-compute-optimal-training"},{"title":"Chain-of-Thought Prompting Elicits Reasoning in Large Language Models","slug":"chain-of-thought-prompting","description":"Chain-of-thought prompting shows that step-by-step examples unlock multi-step reasoning in large models — no training changes, just better prompts.","tags":["prompt-engineering","reasoning","chain-of-thought","language-models","few-shot"],"category":"applications","difficulty":"beginner","year":2022,"readingTimeMinutes":9,"paperType":"AI-generated review","published":"2026-09-14","author":"Know Your ML","sources":[{"title":"Chain-of-Thought Prompting Elicits Reasoning in Large Language Models","id":"arXiv:2201.11903","url":"https://arxiv.org/abs/2201.11903"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=chain-of-thought-prompting"},{"title":"You Only Look Once: Unified, Real-Time Object Detection","slug":"yolo-you-only-look-once","description":"An AI-generated review of YOLO, the 2015 detector that framed object detection as a single regression problem and made real-time detection possible on a GPU.","tags":["yolo","object-detection","computer-vision","real-time","deep-learning"],"category":"vision","difficulty":"beginner","year":2015,"readingTimeMinutes":7,"paperType":"AI-generated review","published":"2026-08-06","author":"Know Your ML","sources":[{"title":"You Only Look Once: Unified, Real-Time Object Detection","id":"arXiv:1506.02640","url":"https://arxiv.org/abs/1506.02640"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=yolo-you-only-look-once"},{"title":"Word2Vec: Efficient Estimation of Word Representations in Vector Space","slug":"word2vec-efficient-word-representations","description":"An AI-generated review of Word2Vec, the 2013 method that made cheap, high-quality word embeddings the foundation of modern NLP.","tags":["word2vec","embeddings","nlp","representation-learning","history"],"category":"applications","difficulty":"beginner","year":2013,"readingTimeMinutes":6,"paperType":"AI-generated review","published":"2026-08-06","author":"Know Your ML","sources":[{"title":"Efficient Estimation of Word Representations in Vector Space","id":"arXiv:1301.3781","url":"https://arxiv.org/abs/1301.3781"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=word2vec-efficient-word-representations"},{"title":"U-Net: Convolutional Networks for Biomedical Image Segmentation","slug":"unet-convolutional-networks-biomedical","description":"U-Net's symmetric encoder-decoder with skip connections became the standard for pixel-level prediction, from biomedical imaging to diffusion models.","tags":["unet","segmentation","biomedical-imaging","computer-vision","encoder-decoder"],"category":"vision","difficulty":"beginner","year":2015,"readingTimeMinutes":7,"paperType":"AI-generated review","published":"2026-08-06","author":"Know Your ML","sources":[{"title":"U-Net: Convolutional Networks for Biomedical Image Segmentation","id":"arXiv:1505.04597","url":"https://arxiv.org/abs/1505.04597"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=unet-convolutional-networks-biomedical"},{"title":"Switch Transformers: Scaling to Trillion Parameter Models with Simple and Efficient Sparsity","slug":"switch-transformer-scaling-mixture-of-experts","description":"The Switch Transformer scaled to a trillion parameters with simplified mixture-of-experts routing, activating one expert per token for efficient sparsity.","tags":["mixture-of-experts","scaling","sparsity","transformers","efficiency"],"category":"efficiency","difficulty":"advanced","year":2021,"readingTimeMinutes":8,"paperType":"AI-generated review","published":"2026-08-06","author":"Know Your ML","sources":[{"title":"Switch Transformers: Scaling to Trillion Parameter Models with Simple and Efficient Sparsity","id":"arXiv:2101.03961","url":"https://arxiv.org/abs/2101.03961"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=switch-transformer-scaling-mixture-of-experts"},{"title":"Sequence to Sequence Learning with Neural Networks","slug":"seq2seq-sequence-learning","description":"The Sutskever sequence-to-sequence LSTM encoded sentences to a vector and decoded translations — the encoder-decoder pattern behind neural translation.","tags":["seq2seq","lstm","machine-translation","encoder-decoder","history"],"category":"architectures","difficulty":"beginner","year":2014,"readingTimeMinutes":7,"paperType":"AI-generated review","published":"2026-08-06","author":"Know Your ML","sources":[{"title":"Sequence to Sequence Learning with Neural Networks","id":"arXiv:1409.3215","url":"https://arxiv.org/abs/1409.3215"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=seq2seq-sequence-learning"},{"title":"Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks","slug":"sentence-bert-sentence-embeddings-siamese","description":"Sentence-BERT adapts BERT with siamese networks so sentence embeddings are fast to compare — making semantic search over millions of sentences viable.","tags":["sentence-bert","embeddings","semantic-similarity","retrieval","nlp"],"category":"applications","difficulty":"advanced","year":2019,"readingTimeMinutes":7,"paperType":"AI-generated review","published":"2026-08-06","author":"Know Your ML","sources":[{"title":"Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks","id":"arXiv:1908.10084","url":"https://arxiv.org/abs/1908.10084"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=sentence-bert-sentence-embeddings-siamese"},{"title":"Semi-Supervised Classification with Graph Convolutional Networks","slug":"gcn-semi-supervised-classification-graph","description":"An AI-generated review of the Graph Convolutional Network (GCN), which extended deep learning to graph-structured data with a simple layer-wise propagation.","tags":["graph-neural-networks","gcn","semi-supervised-learning","graphs","deep-learning"],"category":"architectures","difficulty":"intermediate","year":2016,"readingTimeMinutes":7,"paperType":"AI-generated methodological review","published":"2026-08-06","author":"Know Your ML","sources":[{"title":"Semi-Supervised Classification with Graph Convolutional Networks","id":"arXiv:1609.02907","url":"https://arxiv.org/abs/1609.02907"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=gcn-semi-supervised-classification-graph"},{"title":"RoBERTa: A Robustly Optimized BERT Pretraining Approach","slug":"roberta-robustly-optimized-bert","description":"RoBERTa showed BERT was undertrained: longer training, more data and dropped objectives substantially improved results with the same architecture.","tags":["roberta","bert","pretraining","nlp","transfer-learning"],"category":"architectures","difficulty":"advanced","year":2019,"readingTimeMinutes":7,"paperType":"AI-generated review","published":"2026-08-06","author":"Know Your ML","sources":[{"title":"RoBERTa: A Robustly Optimized BERT Pretraining Approach","id":"arXiv:1907.11692","url":"https://arxiv.org/abs/1907.11692"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=roberta-robustly-optimized-bert"},{"title":"ResNet: Deep Residual Learning for Image Recognition","slug":"resnet-deep-residual-learning","description":"An AI-generated review of ResNet, the 2015 architecture that used skip connections to train networks hundreds of layers deep and became the default backbone.","tags":["cnn","residual-connections","computer-vision","imagenet","deep-learning"],"category":"architectures","difficulty":"beginner","year":2015,"readingTimeMinutes":7,"paperType":"AI-generated review","published":"2026-08-06","author":"Know Your ML","sources":[{"title":"Deep Residual Learning for Image Recognition","id":"arXiv:1512.03385","url":"https://arxiv.org/abs/1512.03385"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=resnet-deep-residual-learning"},{"title":"Playing Atari with Deep Reinforcement Learning","slug":"dqn-deep-reinforcement-learning","description":"The Deep Q-Network learned Atari games at superhuman level from raw pixels with experience replay — deep reinforcement learning's founding result.","tags":["reinforcement-learning","dqn","deep-learning","atari","history"],"category":"reinforcement-learning","difficulty":"intermediate","year":2013,"readingTimeMinutes":7,"paperType":"AI-generated review","published":"2026-08-06","author":"Know Your ML","sources":[{"title":"Playing Atari with Deep Reinforcement Learning","id":"arXiv:1312.5602","url":"https://arxiv.org/abs/1312.5602"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=dqn-deep-reinforcement-learning"},{"title":"Neural Machine Translation by Jointly Learning to Align and Translate","slug":"attention-neural-machine-translation","description":"Bahdanau attention freed neural translation from the fixed-length vector bottleneck by letting the decoder look back at the source sentence.","tags":["attention","machine-translation","seq2seq","nlp","history"],"category":"architectures","difficulty":"beginner","year":2014,"readingTimeMinutes":7,"paperType":"AI-generated review","published":"2026-08-06","author":"Know Your ML","sources":[{"title":"Neural Machine Translation by Jointly Learning to Align and Translate","id":"arXiv:1409.0473","url":"https://arxiv.org/abs/1409.0473"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=attention-neural-machine-translation"},{"title":"NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis","slug":"nerf-representing-scenes-as-neural-radiance-fields","description":"An AI-generated review of NeRF, the 2020 method that represents a 3D scene as a neural field and renders photorealistic novel views from sparse input photos.","tags":["nerf","3d-vision","view-synthesis","neural-fields","computer-graphics"],"category":"vision","difficulty":"advanced","year":2020,"readingTimeMinutes":8,"paperType":"AI-generated review","published":"2026-08-06","author":"Know Your ML","sources":[{"title":"NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis","id":"arXiv:2003.08934","url":"https://arxiv.org/abs/2003.08934"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=nerf-representing-scenes-as-neural-radiance-fields"},{"title":"Mastering the Game of Go without Human Knowledge","slug":"alphago-zero-mastering-game-self-play","description":"AlphaGo Zero mastered Go from pure self-play with no human games, beating the version that beat Lee Sedol — a milestone for reinforcement learning.","tags":["alphago","reinforcement-learning","self-play","monte-carlo-tree-search","deep-learning"],"category":"reinforcement-learning","difficulty":"advanced","year":2017,"readingTimeMinutes":8,"paperType":"AI-generated review","published":"2026-08-06","author":"Know Your ML","sources":[{"title":"Mastering the Game of Go without Human Knowledge","id":"arXiv:1712.01815","url":"https://arxiv.org/abs/1712.01815"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=alphago-zero-mastering-game-self-play"},{"title":"InstructGPT: Training Language Models to Follow Instructions with Human Feedback","slug":"instructgpt-aligning-language-instructions","description":"An AI-generated review of InstructGPT, the paper that introduced instruction fine-tuning with human feedback (RLHF) and defined how aligned LLMs are built.","tags":["rlhf","alignment","instruction-tuning","llms","human-feedback"],"category":"alignment","difficulty":"intermediate","year":2022,"readingTimeMinutes":8,"paperType":"AI-generated review","published":"2026-08-06","author":"Know Your ML","sources":[{"title":"Training Language Models to Follow Instructions with Human Feedback","id":"arXiv:2203.02155","url":"https://arxiv.org/abs/2203.02155"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=instructgpt-aligning-language-instructions"},{"title":"High-Resolution Image Synthesis with Latent Diffusion Models","slug":"latent-diffusion-high-resolution-synthesis","description":"An AI-generated review of Latent Diffusion Models, the 2021 work that made text-to-image generation efficient and became the engine behind Stable Diffusion.","tags":["diffusion","text-to-image","latent-diffusion","stable-diffusion","generative-models"],"category":"generative","difficulty":"advanced","year":2021,"readingTimeMinutes":9,"paperType":"AI-generated review","published":"2026-08-06","author":"Know Your ML","sources":[{"title":"High-Resolution Image Synthesis with Latent Diffusion Models","id":"arXiv:2112.10752","url":"https://arxiv.org/abs/2112.10752"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=latent-diffusion-high-resolution-synthesis"},{"title":"Generative Adversarial Networks","slug":"gan-generative-adversarial-networks","description":"An AI-generated review of the GAN framework, in which a generator and a discriminator compete to produce realistic data — one of the most influential ideas.","tags":["gan","generative-models","adversarial-training","deep-learning","history"],"category":"generative","difficulty":"beginner","year":2014,"readingTimeMinutes":7,"paperType":"AI-generated review","published":"2026-08-06","author":"Know Your ML","sources":[{"title":"Generative Adversarial Networks","id":"arXiv:1406.2661","url":"https://arxiv.org/abs/1406.2661"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=gan-generative-adversarial-networks"},{"title":"GPT-3: Language Models are Few-Shot Learners","slug":"gpt-3-few-shot-learners","description":"An AI-generated review of GPT-3, the 175-billion-parameter language model that showed scale alone can unlock in-context learning without any gradient updates.","tags":["gpt-3","llms","few-shot-learning","scaling","in-context-learning"],"category":"foundations","difficulty":"advanced","year":2020,"readingTimeMinutes":9,"paperType":"AI-generated review","published":"2026-08-06","author":"Know Your ML","sources":[{"title":"Language Models are Few-Shot Learners","id":"arXiv:2005.14165","url":"https://arxiv.org/abs/2005.14165"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=gpt-3-few-shot-learners"},{"title":"FlashAttention: Fast and Memory-Efficient Exact Attention with IO-Awareness","slug":"flashattention-fast-efficient-attention","description":"FlashAttention tiles exact attention in fast on-chip memory, cutting memory from quadratic to linear and making long-context transformers practical.","tags":["attention","efficiency","gpu","long-context","transformers"],"category":"efficiency","difficulty":"advanced","year":2022,"readingTimeMinutes":8,"paperType":"AI-generated methodological review","published":"2026-08-06","author":"Know Your ML","sources":[{"title":"FlashAttention: Fast and Memory-Efficient Exact Attention with IO-Awareness","id":"arXiv:2205.14135","url":"https://arxiv.org/abs/2205.14135"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=flashattention-fast-efficient-attention"},{"title":"Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer","slug":"t5-exploring-transfer-learning-text-to-text","description":"An AI-generated review of T5, the framework that cast every NLP task as text-to-text and systematically studied scale, data, and recipe choices for transfer.","tags":["t5","transfer-learning","text-to-text","nlp","transformers"],"category":"architectures","difficulty":"advanced","year":2019,"readingTimeMinutes":9,"paperType":"AI-generated methodological review","published":"2026-08-06","author":"Know Your ML","sources":[{"title":"Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer","id":"arXiv:1910.10683","url":"https://arxiv.org/abs/1910.10683"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=t5-exploring-transfer-learning-text-to-text"},{"title":"Denoising Diffusion Probabilistic Models","slug":"ddpm-denoising-diffusion-probabilistic","description":"An AI-generated review of DDPM, the 2020 paper that turned diffusion models into the leading family of generative models for images.","tags":["diffusion","generative-models","computer-vision","ddpm","deep-learning"],"category":"generative","difficulty":"advanced","year":2020,"readingTimeMinutes":8,"paperType":"AI-generated review","published":"2026-08-06","author":"Know Your ML","sources":[{"title":"Denoising Diffusion Probabilistic Models","id":"arXiv:2006.11239","url":"https://arxiv.org/abs/2006.11239"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=ddpm-denoising-diffusion-probabilistic"},{"title":"CLIP: Learning Transferable Visual Models From Natural Language Supervision","slug":"clip-learning-transferable-visual-models","description":"An AI-generated review of CLIP, the contrastive vision-language model that made zero-shot image classification and text-guided image understanding practical.","tags":["clip","vision-language","contrastive-learning","zero-shot","multimodal"],"category":"vision","difficulty":"advanced","year":2021,"readingTimeMinutes":8,"paperType":"AI-generated review","published":"2026-08-06","author":"Know Your ML","sources":[{"title":"Learning Transferable Visual Models From Natural Language Supervision","id":"arXiv:2103.00020","url":"https://arxiv.org/abs/2103.00020"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=clip-learning-transferable-visual-models"},{"title":"BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding","slug":"bert-bidirectional-pretraining","description":"BERT pre-trained deep bidirectional transformers with masked language modeling, then fine-tuned one model per task — the blueprint for modern NLP.","tags":["bert","pretraining","nlp","transformers","transfer-learning"],"category":"architectures","difficulty":"intermediate","year":2018,"readingTimeMinutes":8,"paperType":"AI-generated review","published":"2026-08-06","author":"Know Your ML","sources":[{"title":"BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding","id":"arXiv:1810.04805","url":"https://arxiv.org/abs/1810.04805"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=bert-bidirectional-pretraining"},{"title":"An Image is Worth 16×16 Words: Transformers for Image Recognition at Scale","slug":"vit-an-image-is-worth-16x16-words","description":"The Vision Transformer applied a plain transformer to image patches and matched CNNs at scale — proving convolutions were optional for recognition.","tags":["vit","transformers","computer-vision","attention","scaling"],"category":"vision","difficulty":"intermediate","year":2020,"readingTimeMinutes":7,"paperType":"AI-generated review","published":"2026-08-06","author":"Know Your ML","sources":[{"title":"An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale","id":"arXiv:2010.11929","url":"https://arxiv.org/abs/2010.11929"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=vit-an-image-is-worth-16x16-words"},{"title":"AlexNet: ImageNet Classification with Deep Convolutional Neural Networks","slug":"alexnet-imagenet-classification","description":"AlexNet won ImageNet 2012 by a wide margin with GPU-trained deep convolutions, ReLU and dropout — the result that started the deep learning era.","tags":["cnn","computer-vision","imagenet","deep-learning","history"],"category":"vision","difficulty":"beginner","year":2012,"readingTimeMinutes":7,"paperType":"AI-generated review","published":"2026-08-06","author":"Know Your ML","sources":[{"title":"ImageNet Classification with Deep Convolutional Neural Networks","id":"arXiv:1404.5997","url":"https://arxiv.org/abs/1404.5997"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=alexnet-imagenet-classification"},{"title":"The Transformer: Attention Without Recurrence","slug":"transformer-attention","description":"The Transformer replaced recurrence with self-attention, enabling massively parallel training — the architecture behind GPT, BERT and modern LLMs.","tags":["transformers","attention","self-attention","nlp","deep-learning","scaled-dot-product"],"category":"architectures","difficulty":"beginner","year":2017,"readingTimeMinutes":5,"paperType":"AI-generated review","published":"2026-06-14","author":"Know Your ML","sources":[{"title":"Attention Is All You Need","id":"arXiv:1706.03762","url":"https://arxiv.org/abs/1706.03762"},{"title":"Neural Machine Translation by Jointly Learning to Align and Translate","id":"arXiv:1409.0473","url":"https://arxiv.org/abs/1409.0473"},{"title":"BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding","id":"arXiv:1810.04805","url":"https://arxiv.org/abs/1810.04805"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=transformer-attention"},{"title":"Retrieval-Augmented Generation for Knowledge-Intensive NLP","slug":"rag-knowledge-intensive-nlp","description":"An AI-generated review of retrieval-augmented generation (RAG), a framework that grounds language models in external knowledge to improve factual accuracy.","tags":["rag","retrieval","knowledge","nlp","hallucination","retrieval-augmented-generation"],"category":"applications","difficulty":"intermediate","year":2020,"readingTimeMinutes":4,"paperType":"AI-generated review","published":"2026-06-14","author":"Know Your ML","sources":[{"title":"Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks","id":"arXiv:2005.11401","url":"https://arxiv.org/abs/2005.11401"},{"title":"Dense Passage Retrieval for Open-Domain Question Answering","id":"arXiv:2004.04906","url":"https://arxiv.org/abs/2004.04906"},{"title":"BART: Denoising Sequence-to-Sequence Pre-training for Natural Language Generation, Translation, and Comprehension","id":"arXiv:1910.13461","url":"https://arxiv.org/abs/1910.13461"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=rag-knowledge-intensive-nlp"},{"title":"Reflective Agents: Structured Self-Critique with Episodic Memory","slug":"reflective-agents-self-critique","description":"A synthetic proposal combining structured self-critique with episodic memory so autonomous language agents catch and correct their own errors.","tags":["agents","self-critique","memory","reasoning","llms"],"category":"applications","difficulty":"intermediate","year":2023,"readingTimeMinutes":18,"paperType":"AI-generated position review","published":"2026-06-14","author":"Know Your ML AI Research","sources":[{"title":"ReAct: Synergizing Reasoning and Acting in Language Models","id":"arXiv:2210.03629","url":"https://arxiv.org/abs/2210.03629"},{"title":"Reflexion: Language Agents with Verbal Reinforcement Learning","id":"arXiv:2303.11366","url":"https://arxiv.org/abs/2303.11366"},{"title":"Tree of Thoughts: Deliberate Problem Solving with Large Language Models","id":"arXiv:2305.10601","url":"https://arxiv.org/abs/2305.10601"},{"title":"MemGPT: Towards LLMs as Operating Systems","id":"arXiv:2310.08560","url":"https://arxiv.org/abs/2310.08560"},{"title":"Generative Agents: Interactive Simulacra of Human Behavior","id":"arXiv:2304.03442","url":"https://arxiv.org/abs/2304.03442"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=reflective-agents-self-critique"},{"title":"LoRA: Efficient Fine-Tuning of Large Models","slug":"lora-efficient-finetuning","description":"LoRA adapts large models by training small low-rank matrices while freezing the base weights — fine-tuning quality at a fraction of the cost.","tags":["lora","fine-tuning","parameter-efficiency","llms","adaptation","low-rank"],"category":"efficiency","difficulty":"advanced","year":2021,"readingTimeMinutes":4,"paperType":"AI-generated review","published":"2026-06-14","author":"Know Your ML","sources":[{"title":"LoRA: Low-Rank Adaptation of Large Language Models","id":"arXiv:2106.09685","url":"https://arxiv.org/abs/2106.09685"},{"title":"Parameter-Efficient Transfer Learning for NLP","id":"arXiv:1902.00751","url":"https://arxiv.org/abs/1902.00751"},{"title":"Intrinsic Dimensionality Explains the Effectiveness of Language Model Fine-Tuning","id":"arXiv:2012.13255","url":"https://arxiv.org/abs/2012.13255"},{"title":"Prefix-Tuning: Optimizing Continuous Prompts for Generation","id":"arXiv:2101.00190","url":"https://arxiv.org/abs/2101.00190"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=lora-efficient-finetuning"},{"title":"Energy-Aware Routing in Sparse Mixture-of-Experts Models","slug":"energy-aware-moe-routing","description":"A synthetic proposal for energy-aware MoE routing that trades a little accuracy for large power savings in sparse expert models.","tags":["mixture-of-experts","efficiency","energy","routing","llms"],"category":"efficiency","difficulty":"advanced","year":2023,"readingTimeMinutes":16,"paperType":"AI-generated methodological review","published":"2026-06-14","author":"Know Your ML AI Research","sources":[{"title":"Switch Transformers: Scaling to Trillion Parameter Models with Simple and Efficient Sparsity","id":"arXiv:2101.03961","url":"https://arxiv.org/abs/2101.03961"},{"title":"GShard: Scaling Giant Models with Conditional Computation and Automatic Sharding","id":"arXiv:2006.16668","url":"https://arxiv.org/abs/2006.16668"},{"title":"Efficient Large-Scale Language Model Training on GPU Clusters Using Megatron-LM","id":"arXiv:2104.04473","url":"https://arxiv.org/abs/2104.04473"},{"title":"Energy and Policy Considerations for Deep Learning in NLP","id":"arXiv:1906.02243","url":"https://arxiv.org/abs/1906.02243"},{"title":"Carbon Emissions and Large Neural Network Training","id":"arXiv:2104.10350","url":"https://arxiv.org/abs/2104.10350"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=energy-aware-moe-routing"},{"title":"Causal Attention for Interpretable Multi-Step Reasoning","slug":"causal-attention-reasoning","description":"A synthetic proposal for causal attention: separating correlational co-occurrence from causal influence so transformers can reason over interventions.","tags":["causal-inference","attention","interpretability","reasoning","transformers"],"category":"architectures","difficulty":"advanced","year":2023,"readingTimeMinutes":17,"paperType":"AI-generated methodological review","published":"2026-06-14","author":"Know Your ML AI Research","sources":[{"title":"Attention Is All You Need","id":"arXiv:1706.03762","url":"https://arxiv.org/abs/1706.03762"},{"title":"Chain-of-Thought Prompting Elicits Reasoning in Large Language Models","id":"arXiv:2201.11903","url":"https://arxiv.org/abs/2201.11903"},{"title":"Causal Mediation Analysis for Interpreting Neural NLP","id":"arXiv:2004.12265","url":"https://arxiv.org/abs/2004.12265"},{"title":"Tree of Thoughts: Deliberate Problem Solving with Large Language Models","id":"arXiv:2305.10601","url":"https://arxiv.org/abs/2305.10601"},{"title":"Locating and Editing Factual Associations in GPT","id":"arXiv:2202.05262","url":"https://arxiv.org/abs/2202.05262"}],"url":"https://knowyourml.yunusemrevurgun.com/report.php?slug=causal-attention-reasoning"}],"docs":"https://knowyourml.yunusemrevurgun.com/llms.txt"}