Metraipolis › archive › dc_4dd69c81dc30

Timbre: Reliability-aware harmonic consensus for label-free multimodal test-time rexinforcement learning — arXiv:2610.39439

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.39439
License (as archived)
CC0-1.0
Archived
2026-10-31T06:09:18.792Z
SHA-256 of the archived text
bde6bbf122b1b53b3071277307cc52b4089c66be23dcfc6258f11d1263f25075
Stable URL of this page
https://metraipolis.com/documento/dc_4dd69c81dc30

How to cite

Timbre: Reliability-aware harmonic consensus for label-free multimodal test-time rexinforcement learning — arXiv:2610.39439 — arXiv (en), https://arxiv.org/abs/2610.39439 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_4dd69c81dc30 (Metraipolis archive dc_4dd69c81dc30, sha256 bde6bbf122b1b53b3071277307cc52b4089c66be23dcfc6258f11d1263f25075)
BibTeX
@misc{dc4dd69c81dc30,
  title = {Timbre: Reliability-aware harmonic consensus for label-free multimodal test-time rexinforcement learning — arXiv:2610.39439},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.39439},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_4dd69c81dc30; sha256 bde6bbf122b1b53b3071277307cc52b4089c66be23dcfc6258f11d1263f25075},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.39439. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Label-free test-time reinforcement learning adapts vision--language models using rewards derived from their own predictions. However, misinterpreted visual evidence and sensitivity to textual formulation can produce a consistent but incorrect consensus that majority voting reinforces. Cross-view aggregation can expose such errors, but image and text transformations may themselves degrade task-relevant evidence to different extents. We introduce \method, a reliability-aware harmonic consensus framework for multimodal test-time reinforcement learning. \method constructs a $2\times2$ grid combining original and perturbed images with direct-solving and reframing modes, selects pseudo-labels through weighted harmonic aggregation, and adapts modality-level weights using batch-averaged, temporally smoothed shifts in pseudo-label confidence as a heuristic reliability signal. We formulate consensus as a weighted $\chi^2$-barycenter problem whose unique minimizer is the normalized weighted harmonic aggregate, without requiring correct-answer invariance. An exact support--dispersion decomposition characterizes when harmonic aggregation preserves or corrects arithmetic rankings, while a sensitivity analysis establishes how modality reweighting changes pairwise ranking margins, informing interpretation of the heuristic updates. Across six benchmarks and three Qwen3-VL model scales, \method outperforms TTRV in 16 of 18 benchmark--model settings on adaptation instances and 17 of 18 settings on held-out instances under matched estimation and optimization rollout counts. These results demonstrate improved target-task adaptation and transfer to unseen instances without further updates.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno).
Categorie arXiv: Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2610.39439 — 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 2026-10-31T06:07:08.522Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).