Papers › Escaping the Self-Confirmation Trap: An Execute-Distill-Verify Paradigm for Agentic...

Escaping the Self-Confirmation Trap: An Execute-Distill-Verify Paradigm for Agentic Experience Learning

23 Jun 2026arXiv:2606.24428added by Syntology

Shiding Zhu, Yudi Qi, Yajie Wang, Jiaze Li, Chao Song, Yaorui Shi, Yibo Miao, Hanqi Gao, Kai Zhang

Title, abstract, authors and date from arXiv's metadata (CC0); this paper is not in the Papers with Code archive (frozen 2025-07-28).

Experience-driven self-evolution is critical for large language model (LLM) agents to improve through open-world interaction. However, existing experience learning methods mostly rely on single-agent loops, where the same agent executes tasks, summarizes outcomes, and determines memory content. This setup makes agents vulnerable to the Self-Confirmation Trap: wrong-but-self-consistent trajectories are misidentified as successful experience, leading to cumulative errors during retrieval and reuse. To address this issue, we propose EDV, an Execute-Distill-Verify framework for reliable experience learning. In the Execute stage, multiple heterogeneous agents explore the same task space in parallel to generate diverse candidate trajectories. In the Distill stage, a dedicated third-party agent comparatively analyzes these trajectories to produce candidate experiences, reducing executor-centric summarization bias. In the Verify stage, the execution group validates candidates via a consensus mechanism, and only approved experiences are written into shared or private memory. By decoupling the three stages, EDV transforms experience learning from isolated self-reflection into collaborative construction, filtering erroneous and noisy content before memory insertion. We evaluate EDV on three challenging long-horizon benchmarks: tau2-bench, Mind2Web and MMTB. Results show EDV consistently outperforms strong baselines, validating that reliable experience construction is essential for robust agent self-evolution. Our code is available at https://github.com/shidingz/EDV.

PaperPDFCode Syntology ranMethod as code

In Syntology View this paper on Syntology, its page in Syntology's graph. That page adds which of its functions are also present in other papers' code Syntology holds, with dates, and its dated run record as JSON and a README badge. It also lists the repositories linked to the paper, the abstract and the calls for agents.

Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

For agents, on Syntology's MCP service (how to connect):

Code

Syntology 28 code samples harvested from 1 repository linked to this paper: 20 ran (of which 13 constructed an object rather than computing a result; 15 with no instrument failure: 1 honoured, 1 violated, 13 with no contract checked; 5 where Syntology's instrument failed) · 8 unverified. Where Syntology's instrument failed: 5 where our contract draft was wrong; failures of Syntology's instrument, not of the code.

By repository: found in paper text by Syntology: 28 samples from 1 repository, 20 ran (15 with no instrument failure; 5 where Syntology's instrument failed). The run record, sample by sample. “Ran” means executed on a synthesized input, not that the code is correct or reproduces the paper.

Levels written on 2026-09-22 (UTC): the dates Syntology's graph wrote each sample's verification level, not when the code executed.

shidingz/EDV found in paper text by SyntologySyntology: 28 samples, 20 ran (5 instrument-failed) · 8 unverified report

Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

28 samples harvested: 20 ran (of which 13 constructed an object rather than computing a result; 15 with no instrument failure: 1 honoured, 1 violated, 13 with no contract checked; 5 where Syntology's instrument failed) · 8 unverified. “Where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code. Syntology's record for this page last changed ; when this build read Syntology's graph is in the build record. That date is not when the samples ran.

1ran · honoured contract
1ran · violated contract
5ran · our draft was wrong
13ran
8unverified

Licence: 0 of the 28 samples are pointer only, meaning Syntology does not serve that copy's text. This page shows no code text for any sample; each one links to its file in the repository.

Harvested from shidingz/EDV. “Ran” means the sample executed on a synthesized input. It does not mean the output is correct, and nothing here reproduces the paper's results. A line beginning “fixture:” under a sample that ran gives the arguments Syntology's fixture called the sample with, each as name:type/rank/element kind (? where none was recorded), and the type of what the fixture's run returned, exactly as the harness recorded them; the arguments and the returned type were recorded by two separate runs of the same fixture. No output value is shown, and a returned type says nothing about whether the output was right. “Level recorded” gives the UTC date when Syntology's graph wrote the sample's verification level, not when the code executed; where the level was written before Syntology's graph kept that date, the row says “level recorded, date not recorded”. “Honoured” and “violated” refer to a contract Syntology drafted from the code itself. “Our draft was wrong” and “the contract check could not run” are failures of Syntology's instrument, not of the code: the sample ran on Syntology's synthesized input, and then the drafted contract did not describe it, or the check itself raised an error (RAISES), lacked a dependency (DEP_MISSING) or ran out of time (TIMEOUT); hover the label for which. A metamorphic tier, where shown, is evidence of the output's shape, not that the sample implements the method.

Each sample ends with its code_sha256. A sample's code_sha256 is Syntology's identity for that exact code. It is printed in full (16 hexadecimal characters for every sample in this build), and get_code(code_sha256="…") takes it exactly as printed. An agent calls it on Syntology's MCP service (how to connect); click an identity to copy that call.

Repository labels, per sample. official repository: The archive marks this repository official for the paper. named in the paper: The archive records that the paper mentions this repository; it is not marked official. community (archive-listed): In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper. found in paper text by Syntology: Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted. community: Not in the archive's code links for this paper; a community repository Syntology harvested. Samples are grouped by repository: a repository marked official first, then by how directly the repository is tied to the paper, then in the archive's order; never by how many ran. Licence labels name the repository's licence as recorded at harvest. “Pointer only” means Syntology does not serve that copy's text, for one of six reasons: no licence file was found; the licence was not identified; the licence is recorded as permissive but that copy's record is not marked cleared; the licence is outside the permissive list Syntology serves text under (MIT, Apache-2.0, BSD and similar); this paper's copy was not located, so its licence is not recorded; or this paper's copy is cleared, but the stored copy of the same code that Syntology serves is not. Some licences outside that list permit redistribution, such as WTFPL, and GPL-3.0 under its conditions; they are simply not on the list. Hover a licence label for the reason. File links open the file on GitHub at the default branch, which may have changed since the harvest.

A function or class name links to its page when the name occurs in the harvested code of at least 5 papers; the name is a string, not an identity.

shidingz/EDV found in paper text by Syntology 28 samples, 20 ran (5 instrument-failed) · 8 unverified
AbilityMatrix src/edv/pipeline.py ran MIT (permissive) · f2f59313328a59ce · report fixture: returned AbilityMatrix · level recorded 2026-09-22
ContrastiveDistiller src/edv/pipeline.py ran MIT (permissive) · 93c06a1ee5084c61 · report fixture: returned ContrastiveDistiller · level recorded 2026-09-22
Experience in 5 papers src/edv/pipeline.py ran MIT (permissive) · 655a360541b6efd9 · report fixture: returned Experience · level recorded 2026-09-22
MemoryBank in 10 papers src/edv/pipeline.py ran MIT (permissive) · 9079b690f5039ebd · report fixture: returned MemoryBank · level recorded 2026-09-22
MemoryHit src/edv/pipeline.py ran MIT (permissive) · 46b0a722f6123d8b · report fixture: returned MemoryHit · level recorded 2026-09-22
Observation src/edv/pipeline.py ran MIT (permissive) · 75ed7733fb5fc3e8 · report fixture: returned Observation · level recorded 2026-09-22
Step src/edv/pipeline.py ran MIT (permissive) · a49713ef8b941f3a · report fixture: returned Step · level recorded 2026-09-22
Task in 17 papers src/edv/pipeline.py ran MIT (permissive) · 7e6bf386dea53a1e · report fixture: returned Task · level recorded 2026-09-22
ToolCall src/edv/pipeline.py ran MIT (permissive) · b6632fab06bf37ff · report fixture: returned ToolCall · level recorded 2026-09-22
ToyTranslationEnvironment src/edv/pipeline.py ran MIT (permissive) · c65f5d2c7acb3a49 · report fixture: returned ToyTranslationEnvironment · level recorded 2026-09-22
Trajectory in 7 papers src/edv/pipeline.py ran MIT (permissive) · d14951a6ebdfc386 · report fixture: returned Trajectory · level recorded 2026-09-22
VerificationDecision src/edv/pipeline.py ran MIT (permissive) · 571a62bc659e3932 · report fixture: returned VerificationDecision · level recorded 2026-09-22
Vote src/edv/pipeline.py ran MIT (permissive) · c106bc344ca3423b · report fixture: returned Vote · level recorded 2026-09-22
_dedupe_hits src/edv/pipeline.py ran · our draft was wrong MIT (permissive) · 4747943e918596b7 · report fixture: signature hits:list/1/MemoryHit · returned list · level recorded 2026-09-22
_experience_from_dict src/edv/pipeline.py ran · our draft was wrong MIT (permissive) · e5f491fa3700bf6e · report fixture: signature data:dict/0/? · returned Experience · level recorded 2026-09-22
_jaccard src/edv/pipeline.py ran · honoured contract MIT (permissive) · 8e7c99d2b2d8ce78 · report fixture: signature left:set/0/?, right:set/0/? · returned float · level recorded 2026-09-22
_language_code src/edv/pipeline.py ran · our draft was wrong fingerprinted MIT (permissive) · bf8c7497fd576e6c · report fixture: signature value:str/0/str · returned str · level recorded 2026-09-22
_memory_mentions_iso_codes src/edv/pipeline.py ran · violated contract MIT (permissive) · 2af788ecabcf50b5 · report fixture: signature memories:list/1/Experience · returned bool · level recorded 2026-09-22
_task_word src/edv/pipeline.py ran · our draft was wrong MIT (permissive) · 978206f2f397e75c · report fixture: signature task:Task/0/? · returned str · level recorded 2026-09-22
_tokens src/edv/pipeline.py ran · our draft was wrong MIT (permissive) · 549a85e88e7d3a60 · report fixture: signature text:str/0/str · returned set · level recorded 2026-09-22
BaseAgent in 20 papers src/edv/pipeline.py unverified MIT (permissive) · 348c63d8240bb48c · report
ConsensusVerifier src/edv/pipeline.py unverified MIT (permissive) · 746538fbeaa1b75a · report
ConstructionReport src/edv/pipeline.py unverified MIT (permissive) · eb3bd6fbbcdd0877 · report
EDVPipeline src/edv/pipeline.py unverified MIT (permissive) · 2447ee68ca4c6490 · report
InferenceReport src/edv/pipeline.py unverified MIT (permissive) · ea6064f3dc27a6fb · report
IsoCodeAgent src/edv/pipeline.py unverified MIT (permissive) · 505f5a282720fa3f · report
NaturalLanguageAgent src/edv/pipeline.py unverified MIT (permissive) · dccad036cef55961 · report
_stable_id src/edv/pipeline.py unverified MIT (permissive) · 0074accb2a0b87e2 · report

Method as code Syntology

What Syntology's graph records as this paper's claimed method, lifted into code: each assembly names its bricks (functions and classes), each with its own role and its own verification level. An assembly is not a check that the code implements the method correctly or reproduces the paper's results. Levels are each brick's own verification level in Syntology's graph (level 0: no recorded run), listed brick by brick and never combined into one level; “ran” means the brick executed on a synthesized input, not that its output is correct, and the instrument figure counts failures of Syntology's instrument, not of the code. Where a model read the code against the paper's claim, each assembly's label verdict is one model's judgement, kept in this page's JSON twin; it is not calibrated, so no card prints it.

1 assembly on this paper: 1 harvested, 0 generated.

harvested Assembly for EDV (graph method: EDV). 28 bricks; individually on a fixture: 20 ran (of which 13 constructed an object rather than computing a result; 15 with no instrument failure: 1 honoured, 1 violated, 13 with no contract checked; 5 where Syntology's instrument failed) · 8 unverified. The assembly has not been run end-to-end.

  1. EDVPipeline entry · level 0 · unverified · 2447ee68ca4c6490
  2. AbilityMatrix helper · level 1 · ran · f2f59313328a59ce
  3. BaseAgent model · level 0 · unverified · 348c63d8240bb48c
  4. ConsensusVerifier model · level 0 · unverified · 746538fbeaa1b75a
  5. ConstructionReport helper · level 0 · unverified · eb3bd6fbbcdd0877
  6. ContrastiveDistiller model · level 1 · ran · 93c06a1ee5084c61
  7. InferenceReport helper · level 0 · unverified · ea6064f3dc27a6fb
  8. IsoCodeAgent helper · level 0 · unverified · 505f5a282720fa3f
  9. MemoryBank data_transform · level 1 · ran · 9079b690f5039ebd
  10. NaturalLanguageAgent helper · level 0 · unverified · dccad036cef55961
  11. Task helper · level 1 · ran · 7e6bf386dea53a1e
  12. ToyTranslationEnvironment helper · level 1 · ran · c65f5d2c7acb3a49
  13. _dedupe_hits helper · level 1 · ran · our draft was wrong · 4747943e918596b7
  14. Experience helper · level 1 · ran · 655a360541b6efd9
  15. Step helper · level 1 · ran · a49713ef8b941f3a
  16. ToolCall helper · level 1 · ran · b6632fab06bf37ff
  17. Trajectory helper · level 1 · ran · d14951a6ebdfc386
  18. Vote helper · level 1 · ran · c106bc344ca3423b
  19. _stable_id helper · level 0 · unverified · 0074accb2a0b87e2
  20. VerificationDecision helper · level 1 · ran · 571a62bc659e3932
  21. MemoryHit helper · level 1 · ran · 46b0a722f6123d8b
  22. _language_code helper · level 1 · ran · our draft was wrong · bf8c7497fd576e6c
  23. _task_word helper · level 1 · ran · our draft was wrong · 978206f2f397e75c
  24. _experience_from_dict helper · level 1 · ran · our draft was wrong · e5f491fa3700bf6e
  25. _jaccard helper · level 1 · ran · honoured contract · 8e7c99d2b2d8ce78
  26. _tokens helper · level 1 · ran · our draft was wrong · 549a85e88e7d3a60
  27. _memory_mentions_iso_codes helper · level 1 · ran · violated contract · 2af788ecabcf50b5
  28. Observation helper · level 1 · ran · 75ed7733fb5fc3e8

Each brick ends with its code_sha256, Syntology's identity for that exact code; click one to copy the get_code call that returns it. For agents, on Syntology's MCP service: assembly(arxiv_id="2606.24428") returns these assemblies with each brick's role and wiring (how to connect).

Results from the paper

The Papers with Code archive ends with its 2025-07-28 snapshot. This paper's arXiv identifier, 2606.24428, was issued in June 2026, after that date, so the archive has no leaderboard rows for it.

Placed on leaderboards by Syntology Syntology

Syntology's extractor, the model Claude Sonnet 4.5, judged this paper's tables to report on 1 leaderboard (a model's judgement, not a result) and has not placed the paper on it: task not recorded in the archive · Mind2Web (refused by a rule: the row label the model quoted is not the row's label). What is not shown.

Syntology has checked 8,886 of the 18,637 papers on this site that are newer than the archive (for 3,717 of them no archive leaderboard matched the paper's tables, so there was nothing further to check); 733 were read and have nothing to place (arXiv has no HTML version of the paper, or that version has no tables), 41 could not be read (the extractor's reply could not be parsed), and results from the other 8,977 appear after they are checked.

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections