Get Started
Research questionHow can LLM agents reuse execution traces without losing temporal and outcome-dependent behavior?LLM agents often retrieve or summarize past trajectories too shallowly, losing dependencies across steps and the conditions associated with success or failure. Effective reuse must preserve how behaviors unfold over time while distinguishing outcome-relevant patterns from unreliable shortcuts.
AI Agents
AI Memory
Latest papersRecent research connected to this question, newest first.Trace2Tower: Transition-Aware EigenTrace Induction of Multi-Level Skills for LLM AgentsThis concerns LLM agents learning from execution traces in interactive tasks. The evidence covers ALFWorld and WebShop, reporting task success, step efficiency, invalid actions, exact success, and context-efficient experience reuse; it does not establish behavior beyond these benchmarks.research paper · Sep 4, 2026
Related questions
How can we train and evaluate LLM agents for tool use across single- and multi-turn workflows with serial or parallel calls?How can LLM agents reuse procedural skills across diverse long-horizon tasks without generic or instance-bound memory?How can LLM agents jointly adapt reasoning policies and hierarchical skill libraries during reinforcement learning?How can LLM orchestrators preserve continuous state when collaborating with non-language agents?
Home
Topics
Search
Library