Metraipolis › archive › dc_c2ec6036e660
Consolidating Reasoning with Test-Time Training — arXiv:2611.02104
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2611.02104
- License (as archived)
- CC0-1.0
- Archived
- 2026-11-03T11:07:29.664Z
- SHA-256 of the archived text
618cbf8a6fb60debde26c5caf36fdc95c6d1396b9aaef1bd00a14a3ed0c659ce- Stable URL of this page
https://metraipolis.com/documento/dc_c2ec6036e660
How to cite
Consolidating Reasoning with Test-Time Training — arXiv:2611.02104 — arXiv (en), https://arxiv.org/abs/2611.02104 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_c2ec6036e660 (Metraipolis archive dc_c2ec6036e660, sha256 618cbf8a6fb60debde26c5caf36fdc95c6d1396b9aaef1bd00a14a3ed0c659ce)
BibTeX
@misc{dcc2ec6036e660,
title = {Consolidating Reasoning with Test-Time Training — arXiv:2611.02104},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2611.02104},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_c2ec6036e660; sha256 618cbf8a6fb60debde26c5caf36fdc95c6d1396b9aaef1bd00a14a3ed0c659ce},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.02104. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Effective multi-agent collaboration requires finalizing solutions from reasoning that is distributed across the individual attempts of many agents. However, useful intermediate reasoning can remain unexploited when collaboration is limited to voting on conclusions reached among individual agents, or aggregating long, noisy collections of reasoning trajectories in context. In this work, we propose TTT-Consolidate, a framework that uses multi-agent reasoning trajectories as adaptation data to improve a model's subsequent reasoning. TTT-Consolidate jointly consolidates trajectories into parametric representations through test-time training, then generates a new solution by conditioning on the consolidated information. We meta-learn how to consolidate using bi-level optimization at training time, where an inner loop encodes reasoning trajectories into a lightweight LoRA adapter, and an outer loop learns meta-parameters for the inner loop such that the consolidation is helpful for the model to synthesize a final solution. As no large-scale corpus of diverse reasoning trajectories exists for training aggregation methods, we construct the OpenParallelThinking corpus composed of 12,810 math problems paired with a mixture of trajectories from six LLMs. Evaluated on challenging math and STEM benchmarks, TTT-Consolidate outperforms the best aggregation baselines by up to 20 percentage points, even under matched compute and runtime settings, with the gap widening at larger model scales. TTT-Consolidate also recovers from unsuccessful individual trajectories more robustly, an ability that improves as the number of available (even incorrect) trajectories increases. Finally, we demonstrate that TTT-Consolidate benefits from both SFT and RL-based outer loop objectives. We will release TTT-Consolidate and OpenParallelThinking for use in future work on reasoning aggregation. Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno). Categorie arXiv: Computation and Language (cs.CL) ; Artificial Intelligence (cs.AI). Pagina del lavoro: https://arxiv.org/abs/2611.02104 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito). Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina. Letto il 2026-11-03T10:59:46.076Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).