Search: “papers”

Fascinating paper from Anthropic and colleagues. They study whether models can tell when they are being tested. It turns out that capable models can tell when they are being tested rather than deploye

From GPT-4 to GPT-6. Sparks was a remarkably prescient paper that got a lot of pushback at the time, but absolutely sensed where the vibes were heading with LLMs based on a lot of qualitative experime

Paper Instruments benchmarked sandbox providers for their RL rollouts, and E2B's tool execution came in up to 3x faster than every other sandbox provider they tested, reducing idle GPU time and traini

This 'recurrent depth' is essentially what's in Sec. 5.3 of the 2015 paper: On Learning to Think: Algorithmic Information Theory for Novel Combinations of Reinforcement Learning Controllers and Recurr

Neat paper suggesting that human augmentation and task automation are not necessarily related. Models that are really good at doing work are not always good at helping humans do work better. Given the

Very interesting paper on recursive self-improvement. The whole stack is released. Machine learning engineering gives recursive self-improvement a concrete, executable testbed. OpenMLE is an open full

Impressive paper! It's on one of the hardest tasks for coding agents today. Of course, I am talking about kernel optimization. Coding agents are usually not so great at this. Reasons: Unfamiliar low-l

Very interesting paper on LLM reasoning. They find that frontier models can exhibit invisible reasoning by leveraging semantically irrelevant filler tokens. In other words, invisible reasoning can ser

Great technical paper from Harvard and MIT. It's on role drift in compound LLM systems. (bookmark it) End-to-end RL improves the accuracy of a multi-module LLM pipeline without constraining how the mo

New research from NVIDIA. They just dropped a PyTorch-native training framework for agentic RL. (bookmark it) Paper summary: Molt is a PyTorch-native agentic RL framework with an unusual design target

Great paper on self-improving agent harnesses. (bookmark it) If you maintain a production agent harness, finding every file behind one behavior is often harder than writing the edit. Harness Handbook

Are structured outputs in agents always good? This paper suggests that you might have to take a closer look. Your product's structured output surface is measurably more homogeneous than the chat surfa

// Global Workspace in LLMs // arXiv paper for the popular J-space work from Anthropic. (bookmark it) The short recap: If you build on chain-of-thought or steering vectors, this work provides a mechan

NEW paper from Microsoft and colleagues. Debugging agent trajectories at scale is challenging. This is a clever approach to monitor and improve agents in production. The problem: Finding which step in

// Memory becomes an action space // Great paper on long-term memory for agents. (bookmark it) In short, it's discusses the use of a learned policy for using memory at the right granularity. Most memo

Banger compression paper from NVIDIA. (bookmark it) Bigger MoE models keep winning on quality, but serving them at interactive latency is still hard. NVIDIA compresses the hybrid MoE Nemotron-3-Super

// MetaSkill-Evolve // Great paper on self-improving agents. Most self-improving agents rewrite what the agent does and leave how it improves frozen and hand-authored. That improvement procedure never

Excited to share our paper, “Learning Multi-Agent Coordination via Sheaf-ADMM” to be presented at #ICML2026 Blog: https://pub.sakana.ai/sheaf-admm/ Most AI models process information as one giant, mon

NEW paper worth reading. (bookmark it) The basic idea is to pair a compressive recurrent state with a small exact memory, which helps to recover long-range recall without giving up the efficiency of l

Another fascinating paper on LLM Judges. (bookmark it) It's from Amazon, and they show that if you run panels of LLM judges, averaging their scores is a trap. 'Overall, we establish that robust aggreg

NEW paper from NVIDIA. They discuss robot programming that compounds experience instead of throwing it away. Traditional robot programming forces you to orchestrate perception, contact dynamics, diver

Great paper on managing agent skills. Skill libraries keep growing, and picking the right skills has become a bottleneck for coding agents. The defaults are to expose the agent to the whole skill coll

If you build with MCPs, this one is worth reading. (bookmark it) The paper covers five recurring MCP server patterns across fifteen independently developed servers. That taxonomy is useful because I s

// Neural procedural memory // Good paper on agent memory beyond prompt retrieval. NPM stores procedural skills as activation steering vectors distilled from contrastive historical experience. Textual

Fascinating paper on self-improving agents. (bookmark it) If you are working on agentic loops, you will quickly realize that they are only as good as the effectiveness of the evaluator. Self-improveme

NEW paper worth reading. Reasoning-data curation is expensive because scoring a trace usually means reading it to the end. This new work from UCLA shows you may not have to. The quality of a reasoning

NEW paper from NVIDIA. (bookmark it) Speed-of-light performance analysis tells you the theoretical floor of a workload, but teams still derive it by hand and freeze it. SOLAR automates the whole thing

New paper on giving LLM agents experience that improves the weights and stays readable at the same time. Agent-experience methods split into two camps. Externalized natural-language rules stay interpr

What is actually limiting model routers for coding tasks? Most routers treat picking a model as a static, one-off classification. This paper identifies the real bottleneck as information deficit. Simp

🧠Self-Harness: Harnesses that improve themselves New paper on agents shaping their own harnesses to improve over time. Not from LangChain, but builds on top of DeepAgents! Three key steps: 1/ Weakness

This post references the foundational paper establishing Moore's Law, which describes the empirical observation that the number of transistors on integrated circuits doubles approximately every two ye

// Self-play with a pinch of human data // Really cool paper combining human demonstrations and self-play RL. 30 minutes of human data, 2500x less than imitation learning, is enough to make self-play

Students without access to LLMs are 2 to 8 times more creative than students with access. That is the finding of a new paper comparing 2,200 college admissions essays written by humans before ChatGPT

New paper on how AI agents are reshaping knowledge work. This is a nice economic read on where agents actually change knowledge work to meet that gap directly. (bookmark it) It studies agent adoption

Striking paper from Wharton. The big conclusion: AI must increase productivity 2.7x -- and quickly -- or tech companies risk bankruptcy with all that entails for the economy. For context: this is how

one popular theory is that research paper alpha* and lab publishing ~died when researchers realized that instead of fighting with marketing depts they could simply walk out the door and get >$100m for

neuro paper finds that the hippocampus is basically an immutable event log that projects a state in graph form…. just like activegraph 🤔 if you’re into neuro and think we live in a simulation, read on

NitroGen just won CVPR Best Paper Honorable Mention!! We are making strides towards general-purpose embodied agents that master not only the real world physics, but also all possible physics across a

// The Meta-Agent Challenge // How good are current agents at self-improving? This is a great paper covering some of the challenges. They propose the Meta-Agent Challenge (MAC), where they give a codi

Outstanding paper on long-horizon agents. (bookmark it) Similar to humans, how do you make agents persist on a difficult task, and how is that useful? And which models today work well on this? This ne

// The Efficiency Frontier // Cool paper on context management. As agents reuse the same documents and histories across many turns, the cheapest context strategy is not fixed. This work describes a pr

Banger paper from Harvard. AutoScientists drops the central planner entirely. Agents interpret shared experimental data, self-organize around promising directions, evaluate proposals before resource a

In our new paper, EDGE-OPD (EviDence GuidEd On-Policy Distillation), we introduce two key ideas for improving On-Policy Distillation (OPD). First, we use guided rollouts that inject privileged context

// Language Models Need Sleep // Let your agents 'sleep', folks. On a serious note, this is a fascinating paper on getting the most from long-horizon agents. Here is the problem with agents today: Att

Every 'self-evolving agent' paper this year has mutated text: prompts, skill files, workflow graphs, memory schemas. MOSS from USTC & HKUST argues this is the wrong layer. The thing that actually brea

Gated DeltaNet-2 is here. 🚀 🔥 New paper: Gated DeltaNet-2: Decoupling Erase and Write in Linear Attention Gated DeltaNet-2 outperforms KDA and Mamba-3, the latest and best recurrent architectures, hea