July
Papers
Links
- https://github.com/google/longfellow-zk Longfellow ZK
- https://eprint.iacr.org/2024/2010 Anonymous credentials from ECDSA
- https://www.weaverobotics.com/isaac-1 Isaac 1 - Weave Robotics
- https://zk.golf/ zkGolf - competition to build the cheapest circuits.
- https://thinkingmachines.ai/news/learning-to-replicate-expert-judgment-in-financial-tasks/ Learning to Replicate Expert Judgment in Financial Tasks
- https://interconnected.org/home/2026/07/03/factories Factories are just rooms
- https://www.jeremycollerfoundation.org/news-and-insights/press-releases/scientists-win-us100-000-in-2026-coller-dolittle-prize Scaling Laws, Honestly
- https://www.completeskeptic.com/p/scaling-laws-honestly Scientists win US$100,000 in 2026 Coller Dolittle prize for latest step forward in interspecies communication research
- https://www.anthropic.com/research/global-workspace A global workspace in language models
- https://www.ranprieur.com/tech.html TechJudge
- https://www.lesswrong.com/posts/zFJ3ZdQwrTWE9jT5S/a-review-of-anthropic-s-global-workspace-paper A Review of Anthropic's Global Workspace Paper
- https://www.1x.tech/discover/neos-hands NEO's Hands | An API to the Physical World
- https://arxiv.org/abs/2601.18734 Self-Distilled Reasoner: On-Policy Self-Distillation for Large Language Models
- https://arxiv.org/html/2601.19897v1 Self-Distillation Enables Continual Learning
- https://arxiv.org/abs/2601.20802 Reinforcement Learning via Self-Distillation
- https://blog.weineng.me/posts/slowest_add/ Data Access Patterns That Makes Your CPU Really Angry
- https://arxiv.org/abs/2606.04511 SparDA: Sparse Decoupled Attention for Efficient Long-Context LLM Inference
- https://oaklab.ai/ Oak Lab
- https://en.wikipedia.org/wiki/Hymenoptera_training Hymenoptera training
- https://lalitm.com/post/git-history/ The git history command deserves more attention
- https://github.com/lucidrains/x-jepa x-jepa (wip)
- https://oaklab.ai/posts/learning-from-experience-instead-of-curated-datasets Learning from experience instead of curated datasets
- https://developer.nvidia.com/blog/ai-model-co-design-hardware-friendly-llm-design AI Model Co-Design: Hardware-Friendly LLM Design
- https://www.benjamin-bai.com/projects/lemario LeMario: A JEPA Model trained on Super Mario Bros
- https://thinkingmachines.ai/news/introducing-inkling/ Inkling: Our Open-Weights Model
- https://artificialanalysis.ai/evaluations/omniscience AA-Omniscience: Knowledge and Hallucination Benchmark
- https://www.doc.ic.ac.uk/~nuric/posts/rl/understanding-markov-decision-processes/ Understanding Markov Decision Processes (MDPs)
- https://tropes.fyi/tropes-md tropes.md - A single file containing all cataloged AI writing tropes
- https://www.kimi.com/blog/kimi-k3 Kimi K3: Open Frontier Intelligence
- https://dl.acm.org/doi/10.1145/3803784.3809282 Synthetic Gestures: An Evolutionary Sketching Machine
- https://thereader.mitpress.mit.edu/ai-insight-problems-quirks-human-intelligence/ AI Is No Match for the Quirks of Human Intelligence (I think most of this is wrong)
- https://arxiv.org/abs/2606.31700 Diffusing Blame: Task-Dependent Credit Assignment in Biologically Plausible Dual-Stream Networks
- https://x.com/i/status/2077650169265590727 The goldmine business of selling data to frontier labs
- https://workshop.cjpais.com/projects/transcribe-cpp transcribe.cpp
- https://www.sunday.ai/blog/act-2-preview ACT-2 Preview: Generalizing Reliability
- https://www.alphaxiv.org/abs/2607.sigreg-tutorial SIGReg from First Principles
- https://github.com/humanlayer/advanced-context-engineering-for-coding-agents/blob/main/benchmarking-opus-5-on-slop-code-bench.md Benchmarking Opus 5 on SlopCodeBench
- https://blog.comma.ai/driving-bugs/ Bugs that broke driving
- https://github.com/mlabonne/llm-datasets LLM Datasets - Curated list of datasets and tools for post-training.
- https://huggingface.co/blog/agent-intrusion-technical-timeline Anatomy of a Frontier Lab Agent Intrusion: A Technical Timeline of the July 2026 Incident
- https://github.com/MoonshotAI/Kimi-K3/blob/main/k3_tech_report.pdf Kimi K3 - Technical Report
- https://openai.com/index/how-two-settings-tripled-our-arc-agi-3-scores/ How enabling two settings tripled our scores on the ARC-AGI-3 benchmark
- https://deepmind.google/blog/gemini-robotics-2-brings-whole-body-intelligence-to-robots/ Gemini Robotics 2 brings whole body intelligence to robots
- https://john.fun/elevators Elevators - john.fun
- https://x.com/i/status/2082856158520398279 Kimi K3, explained: how forgetting made the biggest open model possible
- https://sebastianraschka.com/llms-from-scratch/ch04/04_gqa/ Grouped-Query Attention (GQA)