Stateless Language Agents: Scaling Long-Horizon Automated Research
Stateless Language Agents beat three research-agent frameworks at up to one billion tokens, using 84% fewer tokens on kernels.
Stateless Language Agents keep durable research state in a harness and rebuild a fresh, role-specific context for every agent call. A stateless Advisor reads summarized evidence and assigns experiments to parallel Workers. Against three recent frameworks, SLA gets the best final result on software engineering, kernel optimization, and algorithm design at budgets up to one billion tokens, matching the strongest kernel baseline with over 84% fewer tokens. Ablations show focused contexts and explicit assignments both contribute, while the Advisor uses less than 0.6% of tokens.
- The harness owns research state; agents receive fresh role-specific context each call.
- An Advisor assigns experiments to parallel Workers and uses under 0.6% of tokens.
- SLA wins every task and matches the best kernel baseline with over 84% fewer tokens.
- Evaluated on software engineering, kernel optimization, and algorithm design up to one billion tokens.
- Products
- Stateless Language Agents
From the original article
1 min read at huggingface.co
Automated research systems increasingly run LLM agents over long horizons, but more inference does not by itself produce more progress: agents replay growing histories, duplicate one another's work, or stop experimenting while token consumption continues. Yet most evaluations use short budgets or benchmarks that saturate early, leaving these failure modes untested.
By Qizheng Zhang, Changxiu Ji, Isaac Sun · huggingface.co
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.