This work proposes ChronoVision, a multimodal framework designed to align visual logic with latent imagery, and introduces Vbvr-VQA, a novel dataset that evaluates temporal tracking by reformulating video reasoning into a strict image-ordering task.
Abstract
Multimodal large language models excel at passive perception but struggle with complex visual cognitive tasks requiring multi-step temporal reasoning. This degradation largely stems from the inherent ambiguity of language-based reasoning, which often fails to accurately articulate continuous visual transformations. To address this, we propose ChronoVision, a multimodal framework designed to align visual logic with latent imagery. During supervised fine-tuning, a Reconstructive Visual Head predicts the latent representation of the final transformed state, while an ROI Attention Locating module focuses the model on key visual evidence via semantic span queries. In post-training, we apply reinforcement learning with an implicit process grounding mechanism, guided by a composite reward function that evaluates outcome correctness, latent process alignment, and unsupervised visual focus. Furthermore, we introduce Vbvr-VQA, a novel dataset that evaluates temporal tracking by reformulating video reasoning into a strict image-ordering task. Experiments demonstrate that ChronoVision achieves state-of-the-art performance on Vbvr-VQA with 74.8% in-domain and 71.6% out-of-domain accuracy, alongside a strong 55.0% accuracy on IntPhys2, a highly challenging cross-domain benchmark.
Latent visual reasoning aims to perform multimodal reasoning through hidden-state computation rather than explicit textual chains of thought. However, visual information being present in a latent state does not imply that the model actually relies on that state when producing its answer, especially when alternative image-conditioned paths remain available. We introduce Causal Visual Recurrent Reasoning (CVRR), which preserves pretrained visual competence while making recurrent computation the required image-conditioned path to prediction. CVRR initializes recurrence from the question hidden state after the pretrained vision-language model has incorporated the image, then repeatedly updates this state while re-reading the same fixed visual evidence. Before decoding, visual states and the original multimodal KV cache are removed so that only the final recurrent state carries image-conditioned information to the answer. Across the $V^*$, MMVP, BLINK, and MME-RealWorld-Lite benchmarks, CVRR retains strong performance under this strict interface, while compatible latent reasoners fail to recover comparable visual competence even when retrained under the same constraint. Causal interventions further show that predictions remain sensitive to recurrent content when the question is held fixed, and that persistent visual evidence causally revises the recurrent trajectory. These results distinguish latent informativeness from latent computation that is actually used for prediction.
Suhyeong Park, Junha Jung, Jaewoo Kang· 0 citations
GLaQ, a grounded latent-query framework that replaces sequential latent rollout with a fixed set of context-conditioned queries grounded in the original visual tokens, is introduced, suggesting that direct query-to-image grounding can recover localized evidence from the full image without external visual operations or autoregressive latent rollouts.
Ze-Sheng Yang, Lingling Zhang, Xinyu Zhang et al.· 0 citations
TRAM (TRajectory-derived Auxiliary Memory), a training-free method that augments standard decoding with an auxiliary memory pathway derived from the model's own reasoning trajectory, shows that TRAM improves performance over vanilla decoding on mathematical, scientific, and general visual reasoning tasks without additional training.
Kang Liu, Zijing Wang, Yongkang Liu et al.· 0 citations
Latent reasoning has advanced multimodal reasoning through a two-stage training paradigm: (1) a helper image is encoded into latent tokens to teach visual chain-of-thought during a supervised fine-tuning (SFT) stage, and (2) these latent tokens are further refined with reward feedback during a reinforcement learning (RL) stage. In this paper, we identify two key limitations of this framework, one in each stage. First, the SFT stage typically relies on an off-the-shelf vision encoder to encode the helper image, yielding suboptimal latent representations that may not be well aligned with the downstream reasoning task. Second, existing RL methods treat the latent component only through deterministic regularization, which constrains policy drift but does not create alternative latent trajectories for exploration. To address these limitations, we propose Scaffolding Minds. Our approach learns a dedicated scaffolding encoder that provides an optimized target in latent space, and learns both the mean and variance of the RL sampler. We further show that these two improvements are complementary, together yielding substantial gains over strong baselines. Empirically, our method improves over the strongest latent-reasoning baseline by +9.5% on FrozenLake spatial planning, with the gain widening to +19% at 32x32 grid map, and by +5.2% on average across nine visual-centric reasoning benchmarks.
Haoqiang Kang, Yin-Ping Chen, Lu-Yang Liu et al.· 0 citations
Latent-OPD is proposed, which augments OPD with trajectory-level latent distillation and introduces a progressive teacher-lookahead strategy, which aligns middle-to-late student layers with increasingly deeper teacher layers, establishing Latent-OPD as a highly effective approach to frame-efficient video reasoning.
Aoni Shen, Yongheng Zhang, Yinghui Li et al.· 1 citation
Experiments on perception-intensive visual reasoning benchmarks show that LUT outperforms previous latent reasoning methods and remains competitive with latent-text interleaved methods with lower annotation cost.
J. Kang, Siyue Chen, Ming-Da Li et al.· 1 citation
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
