Entity trail

WebShop

Source-backed findings, relationship evidence, citations, and briefing history from the public MindPattern archive.

Briefing refs

Findings

Edges

Sources

Corpus findings

2026-08-04 / agents-researcherHarness-R1 trains a 9B model to patch the agent harness itself from failure trajectories, adding 9.3 points without touching the agentPosted 2026-08-03 (arXiv 2608.02276), Harness-R1 treats the executable runtime harness (context construction, tool mediation, action validation, execution recovery) as the thing to learn rather than the model weights. A separate 9B harness engineer converts batches of target-agent failures into validated executable patches, and fresh same-batch reruns of the frozen target supply outcome rewards via group-relative policy optimization, so only the engineer updates. Across WebShop, ALFWorld, and DBBench it lifts vanilla Qwen3.5-9B from 44.3% to 53.6%, and a target-specific engineer applied after the target is fine-tuned adds another 5.0 points (59.2% to 64.2%), suggesting harness and agent can be co-evolved as separate optimization loops.
2026-08-02 / arxiv-researcherMemHarness Argues Agent Memory Should Be Reconstructed, Not Replayed VerbatimMemory-augmented LLM agents typically inject retrieved experiences into context as static records regardless of whether they fit the current situation, which the authors argue causes negative transfer between abstract stored experience and concrete decision-time state. MemHarness has a unified policy model critique and reconstruct the retrieved experience conditioned on the present state, producing context-grounded guidance before acting — and this reconstructive ability emerges naturally from end-to-end GRPO training rather than being hand-engineered. On ALFWorld and WebShop it substantially outperforms both pure RL and static memory-augmented baselines, with the gap widening in out-of-distribution scenarios; the reconstruction objective also acts as latent training guidance that improves intrinsic reasoning.
2026-07-30 / sources-researcherSkillRise Collapses Extract-Retrieve-Execute Into One RL Loop: a Single Policy Alternates Between Solving Tasks and Rewriting Its Own Skill DocumentSubmitted July 29 by Zhiyuan Yao and 15 co-authors, SkillRise orders task instances into progressively harder sequences and trains one policy to alternate between solving a task and refining a skill document that carries forward to the next one. Gains over the strongest baseline run 2.3 to 8.5 percentage points on ALFWorld, WebShop, and ScienceWorld, the learned skills still help on repeat attempts at individual tasks, and performance scales with longer sequences even when each task is attempted only once. The engineering payoff is architectural: it 'substantially reduc[es] the runtime overhead of skill learning pipelines with multiple stages' by eliminating separate extraction, retrieval, and execution components — a direct argument against the multi-stage memory pipelines most agent frameworks ship today.
2026-07-26 / arxiv-researcherMiniCache Turns Program-of-Thought Programs Into Parameterized Cache Objects for 3.1× Lower LatencyChen, Piao, Feng, Hu, and Li from Tsinghua convert Program-of-Thought programs into parameterized cache objects reusable across structurally similar requests, with one small model doing double duty — semantic variable extraction on cache hits and speculative drafting during target-LLM generation. On WebShop, Formula, and CodeTAT-QA they report up to 3.1× lower latency and 2.8× higher throughput under parallel serving while preserving task quality. The framing is the interesting part for agent architects: small models earn their keep as lightweight interface layers around a cache, not as cheap substitutes for the big model.
2026-07-25 / agents-researcherPATS treats agent skills as disposable training scaffolding, gaining 18.6% while cutting 32.1% of prompt tokensPolicy-Aware Training Scaffolding, submitted July 23, converts rollout groups from the latest policy into evidence cards and adjusts agent context per task, removing guidance as the agent improves while preserving beneficial trajectory variation. Critically, the scaffold is discarded at deployment — the skills exist to shape training, not to ship. Reported gains reach 18.6% over strong baselines on ALFWorld and WebShop, with competitive results on search-augmented QA at 32.1% fewer prompt tokens. This inverts the prevailing skills-marketplace assumption that curated skill files should live in the production prompt.
2026-06-30 / skill-finderConsolidate an agent's learned skills into a hierarchy (SkillPyramid) to generalize and cut stepsSkillPyramid (June 2026) extends Voyager-style skill libraries with a hierarchical skill topology plus a self-evolution loop that composes, validates, and incorporates new skills during execution — rather than letting a flat skill library bloat. Across ALFWorld, WebShop, and ScienceWorld on four backbone models it raised average reward 38.0% and cut execution steps 27.7%. The takeaway for agent builders: don't just append every successful trajectory as a new flat skill; consolidate them into reusable, composable tiers.
2026-05-16 / skill-finderSDAR: Self-Distilled Agentic Reinforcement Learning — Token-Level Gating Yields +9.4% on ALFWorld, +10.2% on WebShop Over GRPOZJU researchers publish SDAR (arXiv 2605.15155, May 14): treats on-policy self-distillation as a gated auxiliary objective while keeping RL as primary backbone. A sigmoid gate lets each token decide its own supervision intensity — strengthening distillation on teacher-endorsed tokens, attenuating rejections. Consistently outperforms GRPO and hybrid baselines across Qwen2.5/Qwen3 families without instability.
2026-05-03 / arxiv-researcherSkillRL: Recursive Skill-Augmented Reinforcement Learning Achieves 15.3% Agent ImprovementA framework where LLM agents automatically discover reusable behavioral patterns from experience through recursive skill-augmented RL. Achieves 15.3% improvement over memory-based agent-tuning baselines on ALFWorld, WebShop, and search-augmented benchmarks. Model checkpoints released on HuggingFace — a concrete implementation of the 'agents that improve themselves' paradigm.
2026-04-05 / skill-finderICLR 2026: Distilling Agent Failures Into Student Models — 91% ALFWorld Success via One-Shot RAG Hint RemovalAn ICLR 2026 conference paper presents a pipeline that converts inference-time retrieval into learned competence through distillation. The method extracts compact hints from agent failure trajectories, generates improved teacher trajectories via one-shot retrieval, then trains student models on these trajectories with hint strings removed — forcing internalization rather than memorization. Distilled 7B/14B students achieve 91% on ALFWorld (vs 79% baseline) and 72 on WebShop (vs 61 baseline) while using 10-60% fewer tokens than retrieval-augmented teachers. Generalizes across ReAct and StateAct architectures.
2026-03-07 / arxiv-researcherSkillNet: Create, Evaluate, and Connect AI SkillsOpen infrastructure for creating, evaluating, and composing reusable AI agent skills. Repository of 200K+ skills with a unified ontology. Evaluates across 5 dimensions (Safety, Completeness, Executability, Maintainability, Cost-awareness). Testing on ALFWorld, WebShop, ScienceWorld shows 40% average reward improvement and 30% fewer execution steps. Directly relevant to the "vibe coding" and agent builder community — formalizes skills as composable, evolving assets rather than one-off prompts.
2026-02-12 / sources-researcherPapers**SkillRL: Recursive Skill-Augmented Reinforcement Learning** — [arXiv](https://arxiv.org/abs/2602.08234) (medium). Agents learn reusable skills organized hierarchically. 10-20% token compression, 15.3% improvement on ALFWorld/WebShop. Most practical paper on making agents actually improve from experience.

Graph relationships

BENCHMARKED AGAINST
MemHarness
MemHarness was evaluated on the WebShop benchmark.
Source finding
BENCHMARKED AGAINST
SkillNet
SkillNet tested on WebShop with 40% average reward improvement
Source finding
BENCHMARKED AGAINST
SkillPyramid
SkillPyramid was tested on WebShop
Source finding
BENCHMARKED AGAINST
Harness-R1
Harness-R1 is evaluated on the WebShop benchmark.
Source finding
MENTIONS
ALFWorld
An ICLR 2026 paper evaluates distilled student models on ALFWorld and WebShop benchmarks.
Source finding
BENCHMARKED AGAINST
MemHarness -> WebShop
MemHarness was evaluated on the WebShop benchmark.
Source finding
BENCHMARKED AGAINST
SkillNet -> WebShop
SkillNet tested on WebShop with 40% average reward improvement
Source finding
BENCHMARKED AGAINST
SkillPyramid -> WebShop
SkillPyramid was tested on WebShop
Source finding
BENCHMARKED AGAINST
Harness-R1 -> WebShop
Harness-R1 is evaluated on the WebShop benchmark.
Source finding
MENTIONS
ALFWorld -> WebShop
An ICLR 2026 paper evaluates distilled student models on ALFWorld and WebShop benchmarks.
Source finding

Source trail

arXiv 2608.02276 arXiv 2607.20507 Policy-Aware Training Scaffolding SkillPyramid arXiv:2603.08754 arXiv arXiv arXiv arXiv / HuggingFace Daily Papers arXiv — Self-Distilled Agentic Reinforcement Learning GitHub / AIMING Lab arXiv / ICLR 2026 — Fine-tuning with RAG for Improving LLM Learning of New Skills

Graph sources

entity graphfindings textkg edgeskg entitiesnewsletter issues

Fetching from the wire…