Notable data and AI papers

Papers627Labs45

627 papers · 45 labs

Topics, by number of papers

Agents and evaluation147 papersAgents, harnesses, benchmarks and how to measure them.114 in the last 12 monthslatest Nov 2026Open →Multimodal and robotics122 papersVision, video, audio, world models and robot control.103 in the last 12 monthslatest Nov 2026Open →Reasoning81 papersStep-by-step thinking, reinforcement learning for reasoning, math and logic.59 in the last 12 monthslatest Oct 2026Open →Inference and efficiency72 papersServing, attention kernels, KV caches, quantisation: the cost of running models.45 in the last 12 monthslatest Oct 2026Open →Foundation models59 papersTechnical reports for new model families.35 in the last 12 monthslatest Oct 2026Open →Training and scaling48 papersScaling laws, fine-tuning, distributed training and optimisation.31 in the last 12 monthslatest Oct 2026Open →Alignment and safety32 papersMaking models follow intent and resist misuse: RLHF, preferences, jailbreaks, deception, hallucination.17 in the last 12 monthslatest Oct 2026Open →Retrieval and data29 papersEmbeddings, RAG, search over documents and the data models learn from.15 in the last 12 monthslatest Oct 2026Open →Architectures23 papersThe shapes of models: Transformers, state space models, mixture of experts.12 in the last 12 monthslatest Oct 2026Open →Applied AI14 papersModels put to work in a domain: education, science, medicine, code.14 in the last 12 monthslatest Oct 2026Open →