ArticleSiru Ouyang, Jun Yan, I-Hung Hsu, et al.Research
Enabling autonomous agents to evolve directly from operational experience addresses a major barrier to deploying reliable, persistent AI systems in real-world environments.
The authors introduce a compelling synergy between test-time scaling and agent memory. Rather than simply scaling search compute at inference or saving verbatim execution histories, extracting high-level reasoning strategies from contrasts between successes and failures allows agents to demonstrate sustained emergent self-improvement.
Summary generated by AI (Gemini 3.8 Flash, 27 September 2026). It may contain errors – the original source is authoritative.
Similar in content, found by the AI search.
Article:Liyan Tang, Cyrus Rashtchian, Chun-Sung Ferng, Andrew Tomkins, Da-Cheng Juan, Tu Vu
This research paper introduces WikiSkill, a framework that co-evolves AI agent skills alongside a persistent knowledge base (wiki). By disentangling raw execution experience, accumulated knowledge, and executable skills, WikiSkill enables systematic reuse of insights across iterations.
Article:Shuai Shao, Kangning Zhang, Qingyao Li, Shijian Wang, Hao Wang, Wenxiang Jiao, Yuan Lu, Yi Guo, Weiwen Liu, Weinan Zhang
Researchers introduce Harness-R1, a framework that trains a dedicated agent to adapt and patch the runtime harness of LLM agents based on failure logs. Using online reinforcement learning, a separate 9B model learns to generate executable patches for context handling, tool mediation, and error recovery to directly improve task completion.
Article:Hongjun Liu, Yifei Ming, Shafiq Joty, Chen Zhao
A research team introduces HASP (Harnessing LLM Agents with Skill Programs), a framework that converts past agent experiences into executable Program Functions (PFs). Instead of relying on passive textual prompts, PFs act as executable guardrails that trigger on failure-prone states to alter the next action or supply corrective context.
Repository:vectorize-io/hindsight
Hindsight is an MIT-licensed memory system designed to help AI agents learn over time. Rather than merely recalling conversation histories, it aims to overcome the limitations of conventional RAG and knowledge graphs on long-term memory tasks.
41.3kstarsPython
Article:Devansh
Devansh explores 'Stateful Swarms', an architecture tackling volatile contexts and repetitive document processing through a persistent blackboard state. Instead of repeatedly processing large inputs or losing signal across conversational handoffs, specialized agents populate an append-only, structured memory. Evaluated on the Harvey Legal Agent Benchmark, this setup yielded higher task pass rates at significantly lower operational costs.
X post:DAIR.AI
A Google research paper presents an architecture for evolving agent skill libraries by separating raw execution traces, a persistent knowledge wiki, and executable skills.
8.6kbookmarks500.7kviews
Parts of this content were created with the assistance of AI.
