New paper from Microsoft and colleagues on generating enterprise data for agents. If you train or evaluate agents on long-running business work, it's worth your time. (bookmark it) MiniCorp is an office simulator. A firm of agents runs an e-commerce company, discusses decisions in messages and meetings, and sets prices, inventory, and ad spend. A simulated market with customers and adaptive competitors responds, and the firm reacts to the results. The market is checked against published studies of real ones. A four-week 20% discount lifts units by 71%, and sales stay 31% above control after prices return to normal. Checkpoints let the same market state replay with a different decision, giving paired outcomes that real company records cannot provide. Paper: Chat with Paper:

Nice paper on agent memory. It explores an effective routing mechanism for memory in agents. The overall finding is that you want a separate retrieval strategy for each type of memory instead of one embedding search over everything. The authors prove that any single retrieval strategy has an upper bound on expected precision when the memory store holds several types. MemoType classifies each memory and each query by type, retrieves memories that match the query type, and applies the strategy for that type. A new dataset, TriMEM, provides the type labels needed to train the classifier. Recall@1 improves by up to 16.18% across three datasets. Paper:

Banger paper from Meta Superintelligence Labs on user simulators for agent training. Agent RL setups usually let an assistant LLM play the user, so the simulated user is too cooperative and too explicit. A fixed GPT-5.5 agent finds tau-bench tasks easier with these users than with real people. This work trains MIMESIS, a 9B user simulator, on human conversations and 13 behavior patterns observed in real users. It beats Claude Opus 5 on behavioral fidelity by 13.4 points. Agents trained against it outperform agents trained against GPT-5.5 under all nine user simulators they never saw. They find that adding a coaching step that turns the simulator's private reasoning into feedback adds further gains. Paper:

Super interesting NVIDIA paper on choosing base models for coding agents. It's actually a clever way to rank base checkpoints by how well each one is likely to do as a coding agent after post-training. They document that base models are really hard to evaluate on agentic coding tasks. They ran six base models on SWE-bench Verified, and five of them solved zero tasks. So instead, they look at the one step in a coding run that actually fixes the task. In other words, they take tasks that a strong post-trained agent already solved, replay its code edits one by one, and run the tests after each edit. The first edit that makes the tests pass is the decisive edit. Then they give the base model everything that happened before that edit and check whether it can come up with that fix. They score this in three ways. They check how likely the base model is to write the fix, whether it can pick the fix out of a set of rejected patches, and whether any fix it writes on its own passes the tests. All three rankings closely match post-trained SWE-bench Verified scores across ten base and post-trained model pairs. Why is this useful? If you pick checkpoints for agentic post-training, this method can give you a signal before you spend the training budget. Paper:

Interesting new paper from Google DeepMind. Studies agents that follow bad advice from users. Users often suggest a fix that sounds right and is wrong. XYEval adds one confident, misleading hint to tasks from tau2-bench, SWE-bench, Terminal-Bench, HLE and MCP-Atlas. The task and its correct solution stay the same, so any drop comes from the agent taking the advice. Scores fall by up to 46.7% relative across Gemini, Claude Opus 4.8 and GPT 5.5. The drops are larger on easier benchmarks, which suggests more capable models will not fix this on their own. Agents often disagree with the hint in their reasoning, then follow it anyway without telling the user. Compliance shows up almost only in failed runs. A system prompt that warns about the XY problem helps on single-turn tasks and leaves large drops on multi-turn ones like tau2-bench and SWE-bench Verified. Paper:

Banger paper introducing Jev-as-a-Judge. The overall finding is that you want to use a cheap judge for most of your evals and send only the uncertain calls to a frontier model. This paper measures how well that works with JEV, TypeSafe AI's decision-only judge. On 510 held-out preference pairs, a cascade that accepted JEV's confident verdicts and escalated the rest to GPT-6 Astra kept 99% of GPT-6's accuracy at about 57% of its fee. JEV returns a verdict and label probabilities with no reasoning text. It costs $0.044 per 1,000 judgments at a median latency of 0.152 seconds, against $12.182 and 1.885 seconds for GPT-6, about 277 times cheaper. On ordinary preference and evidence-grounded factuality it stays within 3 points of GPT-6 (92.2% against 93.5% on RewardBench, 87.5% against 86.7% on HaluEval). The gap grows to 9 to 20 points on tasks that require checking a derivation or rejecting an elaborately written wrong answer, such as JudgeBench (78.6% against 93.1%). On several benchmarks, JEV's gap to GPT-6 is concentrated in its low-confidence decisions, which is why the cascade works. The escalation threshold did not transfer for every fallback model, so the authors recommend setting it on your own data. Paper:

Impressive paper from Salesforce. It discusses the importance of good verifiers for RL environments. Only 35.8% of the environments in the cleanest public RL collection for terminal agents passed Salesforce AI Research's audit. More details below: With the budget held at 3.5K environments, River-8B averaged 19.4 across four terminal benchmarks, against 17.7 for RL on 3.5K environments sampled at random from the same collection. The audit found reward errors in both directions. Some environments give reward 1 for copying a leaked answer or passing a weak verifier without doing the task. Others give reward 0 to a correct solution because the reference answer or oracle is wrong. Two other public collections were only 10.1% and 3.3% clean. The authors argue that RL mainly shapes behaviors, such as inspecting before acting, verifying before finishing and dropping an approach that keeps failing. Those behaviors reuse skills the model already learned in pre-training and SFT. Their recipe, RIVER, filters defective environments and penalizes turns that repeat an earlier command with nearly the same output. River-8B is the best of the open RL-trained 8B models they evaluated on all four benchmarks. Across models from 2B to 27B, using fewer than 30% of TMax's environments, RIVER increases RL gains by 106% on Terminal-Bench-Lite and 30% on Terminal-Bench v2.1. Paper:

Exciting work from NVIDIA. (bookmark it) Interesting to see this approach to turn public Agent Skills into RL environments. Lots of excitement around RL environments so this is a great read. Skill2Env compiles each Skill into executable terminal tasks. A Codex planner reads the SKILL.md bundle, researches related public assets and splits the Skill into workflows. A Codex creator then builds each task with programmatic tests and a behavioral rubric taken from the Skill's own quality criteria. From about 3.4k crawled Skills, the pipeline produced 7,971 tasks across 13 domains, with software engineering under a quarter of the corpus. Generating them with GPT-5.6 Sol cost over $90k in API usage. After 300 steps of outcome-only RL, Qwen3.8-27B improved from 49.4% to 54.1% on Terminal-Bench 2.1 and from 33.4% to 37.7% pass@1 on S2EBench, their hand-verified held-out benchmark. Adding the rubric to the reward gave smaller benchmark gains, 50.1% on Terminal-Bench 2.1. Given the source SKILL.md, a judge preferred the rubric-trained model's trajectories over the base model's on 73.0% of tasks, against 54.5% for the outcome-only model. Paper: Chat with Paper:

Banger paper from Microsoft Research and colleagues. It studies the potential benefits of agents that share progress while they work. (bookmark it) Communication is still a challenge with multi-agent systems. In this setup, agents have no predefined roles and communicate via a shared directory. They report that a team of k agents that write their findings to a shared directory matches the success rate of 4k agents working independently on ARC-AGI-3. The gap grows with k, and teams reliably solve some tasks that no single agent solves. The same setup beat best@k on polyomino packing and exceeded the prior best-known score. On MNIST compression, a four-agent team wrote a 1,957-byte classifier with 99.4% test accuracy, smaller than the best-known human solution. Independent agents still do better when compute is tight or when there is no clear measure of progress, so the paper also tells you when it might be a good idea to skip communication. Paper: Chat with Paper:

Banger paper from Stanford and Together AI. They show why it might be a good idea to let your agent team learn its own way of working together. (bookmark it) Three models (o3-mini, Claude Sonnet 4 and DeepSeek-V3) averaged 66.7% across five math and physics benchmarks as a self-organizing team. Their strongest member alone scored 48.8%, and a perfect router choosing among the members' independent answers scored 59.0%. On AIME 2026 the team reached 71.2%, 13.4 points above that router. One member reviews the team's earlier exchanges and rewrites the teamwork strategy, covering roles, the order of discussion phases, who participates and how partial answers are combined. The strategies were learned from only 15 AIME 2024 problems and then applied unchanged to held-out AIME 2024 problems and four new benchmarks. Paper:

Interesting work on long-horizon research agents. PrimeScientists can decide where a research agent spends its budget. Achieves 10.3% more reward with 50.6% fewer research attempts, under the same budget. That is PrimeScientist against AutoResearch on 12 AI research tasks. They treat deciding where to spend a research agent's budget as part of the agent's job. PrimeScientist keeps an executable plan tree of competing research directions and their outcomes. An adaptive MCTS policy reads the experimental feedback and the remaining budget and chooses whether to explore a new direction or continue a promising one. The gains also hold on systems, code optimization and ML engineering tasks. If your research agent can propose more experiments than you can afford to run, this is a concrete method for choosing among them. Paper:

Interesting paper on agent memory stored as a linked markdown wiki. Lots of great ideas and insights if you work with LLM Wikis. Wikis are useful for agents because each page holds dense text and the links between pages hold structure. WFM is a Wiki Foundation Model trained to use both at once. It turns an LLM Wiki into a graph and retrieves from it with message passing conditioned on the query, so the text of each page and the link structure shape the result together. The team also built a GPU-to-GPU training protocol that trains 10.5x faster, and reports strong results on five agent memory and multi-hop reasoning benchmarks. If your agent's long-term memory is a folder of linked markdown files, WFM is designed for that format. Paper: Chat with Paper:

Great prompt from Anthropic. It helps steer long runs when using Opus 5.5 in Claude Code. Apparently, on long-running tasks, the model sometimes stops to report instead of continuing. I've noticed that this happens a lot with frontier models.

Impressive paper showing how much the harness changes a coding agent's results. Harnesses do play a huge role in what you are getting out of the models. GPT-5.5 was run inside both Claude Code and Codex on the same 1,000 tasks. With a specialized PowerPoint workflow, it improved inside one harness and got worse inside the other. The harness also changed scores when the prompt was identical. ReFigBench asks coding agents to rebuild real arXiv overview figures as editable PowerPoint slides that keep the text, layout and connections. It covers ten configurations across the GPT, Claude, MiMo and MiniMax families, scored by artifact checks, two families of LLM judges and blinded human comparisons. Perception is still the main bottleneck. The specialized workflow removed native connectors in every configuration, yet human judges still preferred its renderings in most matchups. Paper: Chat with Paper:

Own your intelligence stack, folks. You can't scale a company to the frontier by renting intelligence. Custom models, harnesses, and evals are becoming huge assets. And if you're building something new, you might want to consider working on any one of these, or a combination of them.

Interesting work on long-horizon research agents. PrimeScientists can decide where a research agent spends its budget. Achieves 10.3% more reward with 50.6% fewer research attempts, under the same budget. That is PrimeScientist against AutoResearch on 12 AI research tasks. They treat deciding where to spend a research agent's budget as part of the agent's job. PrimeScientist keeps an executable plan tree of competing research directions and their outcomes. An adaptive MCTS policy reads the experimental feedback and the remaining budget and chooses whether to explore a new direction or continue a promising one. The gains also hold on systems, code optimization and ML engineering tasks. If your research agent can propose more experiments than you can afford to run, this is a concrete method for choosing among them. Paper:

Cool paper showing how effective tuning a system prompt for an agent can be. Recommended paper if you tune agent harnesses. This paper presents EvolveTrade, which treats a trading agent's system prompt as its policy. After each trading interval, a separate Policy Agent reads the decision traces and the realized returns and rewrites the prompt. The backbone model stays frozen. Across several market regimes and two backbone models, the evolved agent beats fixed-prompt baselines on Sharpe ratio and cumulative return in most settings. The rewritten prompts also led the agent to run more code-based analysis and to compute signals that fit the current market regime. Paper:

DigitalOcean Managed Agents are here! They work with major harnesses like Claude Code and Codex. 16,000+ tools. Idle agents stop using CPU. Optimized to help builders scale agents in production. Worth checking out.

Recommended if you work on agent memory. Retrieval is the hardest memory problem in most harnesses because agents keep pulling stale context. AML tests this with a coding track of 150 software tasks, each run with relevant history and again with noisy history.