Video reasoning requires models to locate and track question-relevant evidence across frames. While reinforcement learning (RL) with verifiable rewards improves accuracy, it still struggles to achieve reliable spatio-temporal grounding during the reasoning process. Moreover, improving grounding typically relies on scaled training data or inference-time perception tools, which increases annotation cost or computational cost. To address this challenge, we propose VisonCoach, an input-adaptive RL framework that improves spatio-temporal grounding through visual prompting as training-time guidance. During RL training, visual prompts are selectively applied to challenging inputs to amplify question-relevant evidence and suppress distractors. The model then internalizes these improvements through self-distillation, enabling grounded reasoning directly on raw videos without visual prompting at inference. VisonCoach consists of two components: (1) Visual Prompt Selector, which predicts appropriate prompt types conditioned on the video and question, and (2) Spatio-Temporal Reasoner, optimized with RL under visual prompt guidance and object-aware grounding rewards that enforce object identity consistency and multi-region bounding-box overlap. Extensive experiments demonstrate that VisonCoach achieves state-of-the-art performance under comparable settings, across diverse video reasoning, video understanding, and temporal grounding benchmarks (V-STAR, VideoMME, World-Sense, VideoMMMU, PerceptionTest, and Charades-STA), while maintaining a single efficient inference pathway without external tools. Our results show that visual prompting during training improves grounded video reasoning, while self-distillation enables the model to internalize this ability without requiring prompts at inference time.
Figures & tables
Figure 1: Comparison with previous video reasoning approaches and VisionCoach . Comparing with other video reasoning models, VisionCoach leverages visual-prompt-guided RL with self-distillation to internalize improved spatio-temporal grounding behaviors induced by visual guidance. At inference time, it maintains a single forward-pass reasoning while achieving enhanced grounding performance.
Figure 2: Detailed Architecture of VisionCoach . We introduce VisionCoach , a visual-prompt-guided RL framework for training a spatio-temporally grounded reasoner ( Section 3.2 ). The framework includes Visual Prompt Selector ( VP-Selector ) that predicts optimal visual prompts for challenging inputs ( Section 3.3 ), and object-aware spatial grounding rewards that enforce object identity consistency and multiple predicted bounding boxes ( Section 3.4 ).
Figure 3
Model
What
When (Temporal IoU)
Where (Spatial IoU)
Overall
Acc
Chain1
Chain2
Chain1
Chain2
mAM
mLGM
Proprietary Models
Gemini-2-Flash [ 23 ]
53.0
24.5
23.8
4.6
2.2
26.9
35.6
GPT-4o [ 32 ]
60.8
16.7
12.8
6.5
3.0
26.8
38.2
Tool-calling Methods
VideoZoomer [ 17 ]
47.7
11.1
3.7
2.0
0.5
18.8
24.6
Table 2: Performance on the V-STAR benchmark. VisionCoach achieves strong spatio-temporal reasoning across overall dimensions.
Model
VideoMME
WorldSense
VideoMMMU
PerceptionTest
Overall
Long
Overall
Recognition
Overall
Perception
Overall
Proprietary Models
GPT-4o
71.9
-
42.6
-
61.2
66.0
-
Tool-calling Methods
EgoR1
-
64.9
-
-
-
-
-
EgoR1 (w/o RAG)
58.7
50.8
42.0
40.3
34.4
37.6
65.8
Table 3: Performance across different general video understanding benchmarks. * indicates our implementation performance.
Table 5: Zero-shot spatio-temporal grounding performance on HCSTVG.
Table 8
Figure 4: Additional analysis of VisionCoach . We provide analysis including (a) statistics of visual prompting, (b) inference latency, and (c) the effect of visual prompting on spatial grounding reward.
Figure 5: Spatio-temporal attention map. Visual prompting (VP) improves grounding in both temporal and spatial dimensions. The green box indicates the key frame, while the red box highlights the corresponding spatial region. Temporally, VP increases attention on the correct key frame containing the cowboy. Spatially, VP concentrates attention on the region corresponding to the queried visual attributes (e.g., the cowboy wearing specific clothing), while suppressing irrelevant regions.
Appendix figures & tables11 assets
Supplementary material from the paper’s appendix.
Appendix
Figure 6: VP-Selector training data generation pipeline. Given a video-question pair, we construct multiple visual prompting candidates, generate grounded reasoning with proxy reasoners, and derive pseudo-labels for selecting the most effective visual prompt.
Figure 7: Distribution selected visual prompting of VP-Selector training datasets.
Table 13Table 14Table 15
Acc
Where (Chain 1)
Where (Chain 2)
mAM
mLGM
VideoMME
Open-o3-video ( rspamax )
58.4
25.4
4.6
30.3
41.4
61.9
+ Average IoU ( rspaavg )
58.7
28.0
5.0
30.1
41.3
61.9
+ Identity matching ( rspa )
59.9
27.5
5.4
31.3
43.1
62.1
Appendix
Table 16: Ablation of object-aware spatial reward components on V-STAR and VideoMME.
Avg.obj
Avg.box
Open-o3-video ( rspamax )
0.25
0.38
+ Average IoU ( rspaavg )
0.70
1.18
+ Identity matching ( rspa )
0.83
1.40
Appendix
Table 17: Average number of predicted objects and bounding boxes per generated reasoning.
Gating method
Acc
mAM
mLGM
VideoMME
-
59.6
30.4
41.6
61.9
Fixed (darken)
58.3
29.7
40.6
62.4
Fixed (circle)
57.7
29.3
39.9
62.0
Gating (darken, frame#)
59.6
30.6
42.7
62.6
VP-Selector
60.7
31.3
43.1
62.8
Appendix
Table 18: Ablation of visual prompting assignment to hard samples on V-STAR and VideoMME.
Figure 8: Comparison of grounding with Open-o3-video and VisionCoach .
Figure 9: Prompt to proxy reasoner.
Figure 10: Prompt for extracting key object from question and answer pair.
Video reasoning requires models to identify and verify temporally localized evidence within long video sequences. Recent Video Large Language Models (Video-LLMs) have shown promising reasoning abilities when aligned with reinforcement learning, yet existing approaches typically rely on outcome-based rewards that supervise only the final prediction. Such supervision provides limited guidance on how models should discover the relevant temporal evidence during intermediate reasoning. In this work, we propose TimeThink, a reinforcement learning framework that explicitly guides temporal evidence discovery in Video-LLMs. Our key idea is to treat temporal clue steps as the fundamental optimization primitive of video reasoning, where each reasoning step references a candidate time interval in the video. We introduce a step-wise temporal process reward that provides localized credit assignment for these clues and a joint process--outcome optimization objective that balances reasoning fidelity with task correctness. To enable scalable training, we construct TimeThink-RFT-20K, a dataset with automatically derived temporal evidence segments. Extensive experiments across video reasoning, temporal grounding, and general video understanding benchmarks show that TimeThink consistently improves both temporal localization and reasoning performance, achieving state-of-the-art results among open-source video RL models.
Handong Li, Longteng Guo, Zikang Liu +8
School of Artificial Intelligence, University of Chinese Academy of Sciences · Institute of Automation, Chinese Academy of Sciences · OPPO AI Center, OPPO Inc.
Video MLLMs often struggle with fine-grained spatio-temporal reasoning, sometimes generating correct answers based on irrelevant frames or objects. Although outputting spatio-temporal evidence during reasoning is a promising direction, existing RL frameworks typically rely on geometry-only (IoU) rewards, which can be sensitive to boundary perturbations and overlook semantic alignment. To address this, we propose Semantic Evidence Reward (SER), which reformulates spatio-temporal evidence grounding as a constrained verification task. Instead of computing pixel-level overlap, SER uses a referee VLM as a local checker to evaluate model-generated evidence claims across two dimensions: relevance and localization quality, combined with a temporal penalty. This design reduces the reliance on dense box annotations and enables training directly on standard video QA data. On the V-STAR benchmark, SER achieves 49.6% mLGM, improving by 3.0 points over the strong evidence-grounded baseline Open-o3-Video, demonstrating its potential in enhancing both answer accuracy and evidence grounding.
Sheng Xia, Zhengqin Lai, Tianxiang Jiang +4
Nanjing University · Shanghai Innovation Institute · Harbin Institute of Technology, Shenzhen +2
Reinforcement learning has advanced video reasoning in large multi-modal models, yet dominant pipelines either rely on on-policy self-exploration, which plateaus at the model's knowledge boundary, or hybrid replay that mixes policies and demands careful regularization. Dynamic context methods zoom into focused evidence but often require curated pretraining and two-stage tuning, and their context remains bounded by a small model's capability. In contrast, larger models excel at instruction following and multi-modal understanding, can supply richer context to smaller models, and rapidly zoom in on target regions via simple tools. Building on this capability, we introduce an observation-level intervention: a frozen, tool-integrated teacher identifies the missing spatiotemporal dependency and provides a minimal evidence patch (e.g., timestamps, regions etc.) from the original video while the question remains unchanged. The student answers again with the added context, and training updates with a chosen-rollout scheme integrated into Group Relative Policy Optimization (GRPO). We further propose a Robust Improvement Reward (RIR) that aligns optimization with two goals: outcome validity through correct answers and dependency alignment through rationales that reflect the cited evidence. Advantages are group-normalized across the batch, preserving on-policy exploration while directing it along causally meaningful directions with minimal changes to the training stack. Experiments on various related benchmarks show consistent accuracy gains and strong generalization. Code will be available at https://jethrojames.github.io/FFR/.
Haojian Huang, Chuanyu Qin, Yinchuan Li +1
The Hong Kong University of Science and Technology (Guangzhou) · Knowin AI · Institute of Information Engineering, Chinese Academy of Sciences