MemSkill: consolidating episodic memory into executable skills¶

Fidelity: core mechanism. The deterministic mini-suite does not stand in for the paper's full benchmark.

Paper information¶

Field	Details
Paper	MemSkill: consolidating episodic memory into executable skills (arXiv 2602.02474)
Organization	Nanyang Technological University / University of Illinois Urbana-Champaign / University of Illinois Chicago / Tsinghua University
First public release	2026-02-02 (arXiv v1)
Authors' code	Public implementation
Adapter / method	`memskill`
Local reproduction code	`src/auto_research/agent_research/`

Original paper¶

Motivation and method¶

The controller selects memories from past episodes. The designer compiles recurring successful patterns into skills and updates their versions using new feedback.

flowchart LR
 A["Public input or history"] --> B["MemSkill training and execution loop"]
 B --> C["Paper-specific state or objective"]
 C --> D["Matched-budget evaluation"]

Key figure from the paper¶

Figure 2: Core architecture, modules, and connections. From the original paper; copyright remains with the authors.

Core equations¶

\[ m^*=q_\phi(x,M),\quad s=G(\tau,m^*),\quad M,S\leftarrow\operatorname{Consolidate}(M,S,R). \]

Paper results¶

The paper improves success rates and reduces context costs on long-horizon and cross-task agent benchmarks. No production online A/B test is reported; no online figures are inferred.

Local reproduction¶

PlanBench mini-suite, 120 episodes, seed 42: joint success 1.0000, average cost 0.3200, and nonzero method-specific telemetry.

auto-research agent-eval --method memskill --benchmark planbench-mini --episodes 120 --seed 42
auto-research evolve --model agent --dataset planbench-mini --direction "组合 memskill 与已安装论文算子" --generations 2 --population 4

See the fixed metrics.

Reproduction boundaries¶

The local run checks the paper-specific objective, state updates, and fair budgets. It does not reproduce large models, multi-GPU RL, private environments, real websites, or the full benchmark. It validates the mechanism, not the numerical results in the original tables.