Pith Number

pith:IGJCGWIC

pith:2026:IGJCGWICY2S6RYRZPHRBPEMO24

not attested not anchored not stored refs resolved

TeamTR: Trust-Region Fine-Tuning for Multi-Agent LLM Coordination

Bo Liu, Falong Fan, Siao Liu, Yi Xie, Yuanqi Yao, Yue Zhao

Sequential fine-tuning of multi-agent LLM teams incurs a compounding occupancy shift that scales quadratically with agent count, which TeamTR corrects to linear scaling via trust-region resampling and per-agent divergence control.

arxiv:2605.15207 v1 · 2026-05-01 · cs.LG · cs.MA

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

Add to your LaTeX paper

\usepackage{pith}
\pithnumber{IGJCGWICY2S6RYRZPHRBPEMO24}

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

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

We formalize this as the compounding occupancy shift and prove that stale-occupancy evaluation incurs a penalty that scales quadratically with the number of agents. In contrast, intermediate-occupancy evaluation reduces this to linear scaling. We propose TeamTR, a trust-region framework that resamples trajectories after each component update and enforces per-agent divergence control, yielding rigorous per-update and per-stage improvement lower bounds.

C2weakest assumption

The assumption that resampling full team trajectories after every component update remains computationally tractable and that the per-agent divergence control does not introduce new coordination failures not captured by the stated lower bounds.

C3one line summary

TeamTR is a trust-region framework for multi-agent LLM fine-tuning that resamples trajectories after each update to convert quadratic compounding occupancy shift into linear scaling and yields per-update improvement lower bounds.

References

67 extracted · 67 resolved · 18 Pith anchors

[1] Towards a Science of Scaling Agent Systems · doi:10.48550/arxiv.2512.08296

[2] DeepScaleR: Surpassing O1-Preview with a 1.5B Model by Scaling RL , author=. 2025 , howpublished= 2025

[3] Proceedings of the Twentieth European Conference on Computer Systems , pages=

[4] The Llama 3 Herd of Models · arXiv:2407.21783

[5] Qwen2.5 Technical Report , author=. 2025 , eprint= 2025

Formal links

2 machine-checked theorem links

Receipt and verification

First computed	2026-05-20T00:00:46.251805Z
Builder	pith-number-builder-2026-05-17-v1
Signature	Pith Ed25519 (`pith-v1-2026-05`) · public key
Schema	pith-number/v1.0

Canonical hash

4192235902c6a5e8e23979e217918ed71b5a5f35ffb091b607ccc4f948ffbb81

Aliases

arxiv: 2605.15207 · arxiv_version: 2605.15207v1 · doi: 10.48550/arxiv.2605.15207 · pith_short_12: IGJCGWICY2S6 · pith_short_16: IGJCGWICY2S6RYRZ · pith_short_8: IGJCGWIC

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/IGJCGWICY2S6RYRZPHRBPEMO24 \
  | 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: 4192235902c6a5e8e23979e217918ed71b5a5f35ffb091b607ccc4f948ffbb81

Canonical record JSON

{
  "metadata": {
    "abstract_canon_sha256": "3df3b24ab87b145aa54bc53b238ed98dba8014df0acd9cb823c37fadf69d731d",
    "cross_cats_sorted": [
      "cs.MA"
    ],
    "license": "http://creativecommons.org/licenses/by/4.0/",
    "primary_cat": "cs.LG",
    "submitted_at": "2026-05-01T23:42:57Z",
    "title_canon_sha256": "635ae0a9e89bb3314f080c3e66fb3a478baec995b7ee9d505ccc8b4bcfd39440"
  },
  "schema_version": "1.0",
  "source": {
    "id": "2605.15207",
    "kind": "arxiv",
    "version": 1
  }
}