UC Berkeley Paper Induces Forgetting to Scale Research Agents
Deliberate forgetting turns out to be the scaling lever, challenging the default of agents keeping full history and iterating on it. Researchers and agent builders running long-horizon automated tasks can apply the restart-from-best-candidate trick to their own pipelines with the same compute budget. The paper's code link and Chat with Paper tool are available via the post.
A UC Berkeley paper (arXiv:2610.20563) proposes periodically discarding each agent's session and candidate history, then restarting from the best candidate found so far. The method counters the problem of agents anchoring on past attempts, where extra compute budget stops improving results. Across heuristic programming, kernel optimization and model training tasks, it achieves up to 15% gains over parallel agents and up to 39% over a single agent with the same budget.
Brilliant paper on scaling research agents from UC Berkeley. (bookmark it) They propose an unexpected and interesting way to scale automated research with agents. It's common to see research agents keep their full history and iterate on it to solve a task. After a while, each agent anchors on its own past attempts, makes smaller edits, and extra budget stops improving things. So this work induces forgetting in the agents. Periodically, they discard each agent's session and the candidate history, and then restart from the best candidate found so far. Across heuristic programming, kernel optimization, and model training tasks, this gives up to 15% over parallel agents and up to 39% over a single agent with the same budget. arxiv.org/abs/2610.20563 Chat with Paper: academy.dair.ai/papers/scaling… 💬 6 🔄 1 ❤️ 13 👀 2182 📊 9 ⚡
