Deep Thought Alignment: Trajectory-Level Latent Distillation for Video Reasoning

• URL: http://arxiv.org/abs/2608.16316v1
• Date: Mon, 17 Aug 2026 09:23:49 GMT
• Title: Deep Thought Alignment: Trajectory-Level Latent Distillation for Video Reasoning
• Abstract summary: Large Multimodal Models (LMMs) for video reasoning have long been hindered by the high computational cost of processing vast amounts of visual information.<n>We propose Latent-OPD, which augments On-Policy Distillation with trajectory-level latent distillation.<n> Experiments on six video reasoning benchmarks show that Latent-OPD consistently outperforms output-only OPD.
• Score: 42.05846352410443
• License: http://creativecommons.org/licenses/by-nc-sa/4.0/
• Abstract: Large Multimodal Models (LMMs) for video reasoning have long been hindered by the high computational cost of processing vast amounts of visual information. This dilemma motivates the transfer of the reasoning capabilities of large models to smaller, more efficient ones. On-Policy Distillation (OPD) offers a promising solution by matching output-token distributions along student-generated trajectories. However, video reasoning often depends on evidence accumulated across multiple frames. In this context, output-level supervision only captures information expressed through token predictions and does not directly constrain the latent representations formed during reasoning. To address this limitation, we propose Latent-OPD, which augments OPD with trajectory-level latent distillation. Specifically, our method focuses on the position at the end of each trajectory, where hidden states effectively summarize the accumulated visual evidence and reasoning context. Furthermore, we introduce a progressive teacher-lookahead strategy, which aligns middle-to-late student layers with increasingly deeper teacher layers. Experiments on six video reasoning benchmarks show that Latent-OPD consistently outperforms output-only OPD. Notably, the improvements are particularly pronounced in scenarios with limited frames, long videos, or tasks requiring complex evidence aggregation. These results establish Latent-OPD as a highly effective approach to frame-efficient video reasoning.

Related papers

• Video-OPSD: Exploiting Privileged Visual Evidence for On-Policy Self-Distillation in Video Large Language Models [58.45264308304815]
  On-policy self-distillation (OPSD) has recently emerged as an effective post-training paradigm that improves policy optimization through dense token-level supervision from a privileged self-teacher.<n>We present $textbfVideo-OPSD$, an OPSD framework that exploits privileged visual evidence for both self-teacher construction and knowledge transfer.
  arXiv  Detail & Related papers  (2026-08-27T12:51:32Z)
• OPLD: On-Policy Latent Distillation for Multimodal Reasoning [27.842942669819948]
  multimodal Chain-of-leaved (CoT) improves visual reasoning by incorporating auxiliary visual evidence into intermediate reasoning.<n>Existing approaches remain constrained by externally defined reasoning traces and visual operations.<n>We propose OPLD, a framework that transfers the reasoning capability induced by privileged multimodal CoT into latent reasoning representations.
  arXiv  Detail & Related papers  (2026-07-30T12:57:45Z)
• Contrastive On-Policy Distillation [36.47487086937924]
  We propose COPD, a contrastive On-policy Distillation framework.<n> COPD directly encourages the student model to learn more concise and efficient reasoning strategies.
  arXiv  Detail & Related papers  (2026-07-21T12:35:04Z)
• VideoLatent: Video-Language Learning via Latent Self-Forcing [44.01259508852106]
  VideoLatent is a novel MLLM equipped with a latent injection module tailored for video understanding and reasoning.<n>Our model consistently outperforms existing standard and latent MLLMs on general video understanding and complex video reasoning.
  arXiv  Detail & Related papers  (2026-06-22T05:29:30Z)
• InternVideo3: Agentify Foundation Models with Multimodal Contextual Reasoning [61.87294123934038]
  We present InternVideo3, a framework enhancing multimodal understanding in video tasks.<n>MCR treats understanding as a closed-loop process over a shared, evolving context.<n>We show InternVideo3 achieves strong performance on benchmarks like Video-MME, MLVU, and Ego.
  arXiv  Detail & Related papers  (2026-06-10T15:17:08Z)
• VideoZoomer: Reinforcement-Learned Temporal Focusing for Long Video Reasoning [49.35834435935727]
  VideoZoomer is a novel agentic framework that enables MLLMs to control their visual focus during reasoning.<n>Our 7B model delivers diverse and complex reasoning patterns, yielding strong performance across a broad set of long video understanding and reasoning benchmarks.<n>These emergent capabilities allow it to consistently surpass existing open-source models and even rival proprietary systems on challenging tasks.
  arXiv  Detail & Related papers  (2025-12-26T11:43:21Z)
• Video-R2: Reinforcing Consistent and Grounded Reasoning in Multimodal Language Models [56.851611990473174]
  Reasoning over dynamic visual content remains a central challenge for large language models.<n>We propose a reinforcement learning approach that enhances both temporal precision and reasoning consistency.<n>The resulting model, Video R2, achieves consistently higher TAC, VAS, and accuracy across multiple benchmarks.
  arXiv  Detail & Related papers  (2025-11-28T18:59:58Z)
• STEP: Enhancing Video-LLMs' Compositional Reasoning by Spatio-Temporal Graph-guided Self-Training [87.58996020705258]
  Video Large Language Models (Video-LLMs) have recently shown strong derivation in basic video understanding tasks.<n>Video-LLMs struggle with compositional reasoning that requires multi-step explicit-temporal inference across object relations, interactions and events.<n>We propose STEP, a novel graph-guided self-training method that enables VideoLLMs to generate reasoning-rich finetuning data from any raw videos to improve itself.
  arXiv  Detail & Related papers  (2024-11-29T11:54:55Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.