Pith Number

pith:EWJ3PILC

pith:2026:EWJ3PILCTYJ5P75WZHXBIPY6L2

not attested not anchored not stored refs resolved

Stateful Reasoning via Insight Replay

Ang Li, Bin Lei, Caiwen Ding, Jiachen Yang, Xin Eric Wang

Replaying critical insights from earlier in a reasoning trace keeps them accessible and improves accuracy as chains lengthen in large language models.

arxiv:2605.14457 v1 · 2026-05-14 · cs.AI

Open paper page JSON Open Graph Bundle Merged state Verified badge What is a Pith Number?

Add to your LaTeX paper

\usepackage{pith}
\pithnumber{EWJ3PILCTYJ5P75WZHXBIPY6L2}

Prints a linked badge after your title and injects PDF metadata. Compiles on arXiv. Learn more · Embed verified badge

Record completeness

1 Bitcoin timestamp

2 Internet Archive

3 Author claim open · sign in to claim

4 Citations open

5 Replications open

✓ Portable graph bundle live · download bundle · merged state

The bundle contains the canonical record plus signed events. A mirror can host it anywhere and recompute the same current state with the deterministic merge algorithm.

Claims

C1strongest claim

3-round InsightReplay yields accuracy gains across all 24 settings, with an averaged improvement of +1.65 points over standard CoT, and a largest single-setting gain of +9.2 points on R1-Distill-32B's LiveCodeBench v5 subset.

C2weakest assumption

That the model can reliably extract truly critical insights without introducing noise or errors, and that replaying them will restore accessibility without disrupting the ongoing reasoning flow.

C3one line summary

InsightReplay improves LLM accuracy on reasoning benchmarks by extracting and replaying critical insights to maintain their accessibility during extended chain-of-thought generation.

References

37 extracted · 37 resolved · 6 Pith anchors

[1] Chain-of-thought prompting elicits reasoning in large language models.Advances in neural information processing systems, 35:24824–24837 2022

[2] Self-consistency improves chain of thought reasoning in language models 2023

[3] Deepseek-r1 incentivizes reasoning in llms through reinforcement learning.Nature, 645(8081):633–638 2025

[4] Large language models are zero-shot reasoners.Advances in neural information processing systems, 35: 22199–22213 2022

[5] When more is less: Understanding chain-of-thought length in llms 2025

Formal links

2 machine-checked theorem links

Receipt and verification

First computed	2026-05-17T23:39:06.825383Z
Builder	pith-number-builder-2026-05-17-v1
Signature	Pith Ed25519 (`pith-v1-2026-05`) · public key
Schema	pith-number/v1.0

Canonical hash

2593b7a1629e13d7ffb6c9ee143f1e5eba4e57083221e96f0c297a1eea4df5a9

Aliases

arxiv: 2605.14457 · arxiv_version: 2605.14457v1 · doi: 10.48550/arxiv.2605.14457 · pith_short_12: EWJ3PILCTYJ5 · pith_short_16: EWJ3PILCTYJ5P75W · pith_short_8: EWJ3PILC

Agent API

Resolver JSON Graph JSON Events JSON Schema Signing key

Verify this Pith Number yourself

curl -sH 'Accept: application/ld+json' https://pith.science/pith/EWJ3PILCTYJ5P75WZHXBIPY6L2 \
  | jq -c '.canonical_record' \
  | python3 -c "import sys,json,hashlib; b=json.dumps(json.loads(sys.stdin.read()), sort_keys=True, separators=(',',':'), ensure_ascii=False).encode(); print(hashlib.sha256(b).hexdigest())"
# expect: 2593b7a1629e13d7ffb6c9ee143f1e5eba4e57083221e96f0c297a1eea4df5a9

Canonical record JSON

{
  "metadata": {
    "abstract_canon_sha256": "bacb333a818aac7abf9550ff7cd7519b4c32100efb9fae4cc3fed10a0d728184",
    "cross_cats_sorted": [],
    "license": "http://creativecommons.org/licenses/by/4.0/",
    "primary_cat": "cs.AI",
    "submitted_at": "2026-05-14T06:52:59Z",
    "title_canon_sha256": "e9ae494600d475e69e783e533bca6dbe654e76650bf7186efc7b220f4352fd04"
  },
  "schema_version": "1.0",
  "source": {
    "id": "2605.14457",
    "kind": "arxiv",
    "version": 1
  }
}