Pith Number

pith:Y724IUS3

pith:2025:Y724IUS3R43ZC5N6FQA2E34LPF

not attested not anchored not stored refs pending

ThinkPrune: Pruning Long Chain-of-Thought of LLMs via Reinforcement Learning

Bairu Hou, Jacob Andreas, Jiabao Ji, Kaizhi Qian, Shiyu Chang, Yang Zhang, Yujian Liu

Reinforcement learning with token limits can cut LLM chain-of-thought length in half while dropping accuracy by only two percent on math benchmarks.

arxiv:2504.01296 v1 · 2025-04-02 · cs.CL

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

Add to your LaTeX paper

\usepackage{pith}
\pithnumber{Y724IUS3R43ZC5N6FQA2E34LPF}

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

ThinkPrune results in a remarkable performance-length tradeoff -- on the AIME24 dataset, the reasoning length of DeepSeek-R1-Distill-Qwen-1.5B can be reduced by half with only 2% drop in performance.

C2weakest assumption

That the RL objective with the added token-limit penalty will converge to a policy that preserves core reasoning capability rather than learning superficial shortcuts that only work on the training distribution.

C3one line summary

ThinkPrune halves reasoning length on AIME24 for DeepSeek-R1-Distill-Qwen-1.5B with only 2% performance drop by applying iterative RL under token limits.

Formal links

2 machine-checked theorem links

Cited by

25 papers in Pith

Exploring the System 1 Thinking Capability of Large Reasoning Models

DRP: Distilled Reasoning Pruning with Skill-aware Step Decomposition for Efficient Large Reasoning Models

CLORE: Content-Level Optimization for Reasoning Efficiency

Pruning Long Chain-of-Thought of Large Reasoning Models via Small-Scale Preference Optimization

The Landscape of Agentic Reinforcement Learning for LLMs: A Survey

Receipt and verification

First computed	2026-05-17T23:38:14.733060Z
Builder	pith-number-builder-2026-05-17-v1
Signature	Pith Ed25519 (`pith-v1-2026-05`) · public key
Schema	pith-number/v1.0

Canonical hash

c7f5c4525b8f379175be2c01a26f8b797279fcbc3c81d5a12b78a0690de43b95

Aliases

arxiv: 2504.01296 · arxiv_version: 2504.01296v1 · doi: 10.48550/arxiv.2504.01296 · pith_short_12: Y724IUS3R43Z · pith_short_16: Y724IUS3R43ZC5N6 · pith_short_8: Y724IUS3

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/Y724IUS3R43ZC5N6FQA2E34LPF \
  | 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: c7f5c4525b8f379175be2c01a26f8b797279fcbc3c81d5a12b78a0690de43b95

Canonical record JSON

{
  "metadata": {
    "abstract_canon_sha256": "ac003bb650b4416d87af90a5efc31ec072ca7afd6947d4adfcc624afa51623a6",
    "cross_cats_sorted": [],
    "license": "http://creativecommons.org/licenses/by/4.0/",
    "primary_cat": "cs.CL",
    "submitted_at": "2025-04-02T01:59:26Z",
    "title_canon_sha256": "8ac750f31e605e586bf46d52ef3b42890890427fa911befb0f7ce9074a4bcba2"
  },
  "schema_version": "1.0",
  "source": {
    "id": "2504.01296",
    "kind": "arxiv",
    "version": 1
  }
}