Pith. sign in

REVIEW 4 major objections 4 minor 72 references

CAVE claims that augmenting RL training with a boundary-evidence alignment reward turns visual boundary perception into accurate timestamps, beating TimeLens-7B by about 1.9 points averaged over twelve metrics.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

CAVE adds boundary evidence tokens and an attention alignment reward to RL-based video temporal grounding, improving average timestamp accuracy by about 1.9 points over TimeLens-7B.

T0 review reviewed 2026-08-04 challenge →

load-bearing objection A genuinely new attention-reward idea for RL-based VTG, with a nice diagnosis and clean ablations; the main open question is whether the reward really shapes the timestamps or just the added tokens. the 4 major comments →

arxiv 2608.02078 v1 pith:J2ILBEPD submitted 2026-08-03 cs.CL cs.CV

CAVE: Competence-Aware Visual Boundary Evidence Alignment for Video Temporal Grounding

Wei Jia , Zhicong Lu , Yu Chen , Xiang Wang , Shuai Li , Wenqian Lv , Jiayue Cao , Huaxing liu This is my paper
classification cs.CL cs.CV
keywords video temporal groundinglarge vision-language modelsreinforcement learningvisual evidence alignmentboundary evidence tokensreward designevidence-timestamp misalignmentattention supervision
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper argues that reinforcement-learned video temporal grounding models often see the right boundary—their attention peaks at the true start/end frames—yet still emit off-target timestamps. The authors propose CAVE, which adds dedicated and evidence tokens, warms them up with supervised attention, then rewards each token for concentrating on the ground-truth boundary neighborhood while suppressing distracting salient regions during RL. A competence-aware gate scales this evidence reward by how well each rollout group is already localizing. If correct, CAVE closes a gap that pure outcome rewards leave unconstrained: internal boundary evidence gets translated into accurate timestamps. The paper reports CAVE-7B beating the TimeLens-7B baseline by about 1.9 points averaged over twelve metrics.

Core claim

The central claim is that current VTG RL methods suffer from visual evidence–timestamp misalignment: a substantial fraction of incorrectly localized predictions nonetheless show strong model attention around both ground-truth boundaries, so boundary perception is not reliably converted into the numerical start/end tokens. CAVE attacks this by inserting two special tokens after the predicted interval, <Start> and <End>, whose cross-attention over video frames is supervised in a short warm-up to encode boundary-specific semantics. During RL, the Visual Boundary Evidence Alignment Reward compares each token's mean attention inside a similarity-expanded ground-truth window against its attention

What carries the argument

Boundary-specific evidence tokens <Start>/<End> combined with the Visual Boundary Evidence Alignment Reward (VBEAR) and Performance-Aware Gating (PAGE). The tokens are autoregressively generated after the numerical interval, so they summarize the visual evidence associated with the completed temporal decision. VBEAR measures, per boundary, the ratio of the token's mean attention within the ground-truth boundary neighborhood to its mean attention on salient off-boundary frames, and rewards values in (1,2] with saturation. The burden of the argument is carried by this ratio and the gating: it provides dense, boundary-localized visual supervision that standard IoU rewards cannot.

Load-bearing premise

The policy-gradient signal from VBEAR must transfer backward through the whole autoregressive completion to the earlier timestamp tokens; if this indirect credit assignment is too weak, the reported gains would come from the format warm-up or output regularization rather than from evidence alignment.

What would settle it

Train CAVE with the evidence tokens emitted before the numerical timestamps. If VBEAR then yields no improvement over the same warm-up + GRPO baseline, the reward's benefit depends on the current token order's indirect credit path and the evidence-alignment explanation is falsified.

Watch this falsifier. Get emailed when new claim-graph text bears on it.

If this is right

  • If CAVE is right, RL-based VTG can be improved without extra annotation: the evidence tokens are trained on the same ground-truth intervals already used for IoU rewards.
  • The method should generalize to stronger backbones: CAVE-8B outperforms TimeLens-8B on nearly every metric across the three datasets.
  • The ablation results imply that warm-up of boundary semantics is necessary; removing the attention loss L_att consistently degrades all metrics.
  • Because VBEAR evaluates each boundary independently, it prevents one boundary's strong evidence from masking the other's weakness, yielding gains on strict IoU thresholds (0.7).

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The evidence-token ratio ρ_b could be reused as a model-free confidence score at inference, flagging predictions whose boundary evidence is weak; the paper does not propose this.
  • The similarity-guided window expansion and salient off-boundary reference could be adapted to audio or other modalities, where a parallel evidence–output misalignment may exist.
  • A testable extension: use VBEAR as a verifier in a search/reranking step rather than a policy gradient; since it is a scalar reward on attention, it could rank multiple candidate intervals at inference without further training.
  • The smaller gains on Charades-TimeLens hint that the value of evidence alignment depends on annotation precision; datasets with coarser boundaries may need looser window construction. This is our inference, not the paper's claim.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Referee Report

4 major / 4 minor

Summary. The paper proposes CAVE, a method for video temporal grounding with LVLMs. It introduces two special boundary-evidence tokens, <Start> and <End>, generated after the numerical timestamp tokens. A supervised warm-up stage trains these tokens to attend near the ground-truth boundaries, and a subsequent GRPO stage adds a Visual Boundary Evidence Alignment Reward (VBEAR) computed from the evidence tokens' attention ratios. A performance-aware gate (PAGE) scales this reward according to the group mean IoU of the rollouts. Experiments on Charades/ActivityNet/QVHighlights TimeLens benchmarks are reported for 7B and 8B backbones, with ablations, a matched-IoU evidence-consistency analysis, and a case study. The central claim is that CAVE reduces evidence--timestamp misalignment and improves temporal grounding, with an average gain of about 1.9 points over TimeLens-7B across twelve metrics.

Significance. The paper addresses a genuine and under-explored issue: even when a VTG model's internal visual attention peaks near the ground-truth boundaries, the numeric timestamp output can still be off. Introducing dedicated evidence tokens and rewarding their boundary-focused attention is a reasonable mechanism to inject boundary-level supervision into RL. The matched-IoU analysis in Figure 4 is a good attempt to separate evidence alignment from localization accuracy, and the ablation study isolates VBEAR and PAGE. If the mechanism is confirmed, the contribution would be useful to the RL-based VTG community. The evaluation uses public benchmarks and compares directly with the TimeLens baseline under the same backbone and data, which is a strength. However, the current evidence has several load-bearing gaps, detailed below, that prevent the central claim from being fully established.

major comments (4)
  1. [Method (Eq. 5-8) and Analyses (Figure 4)] VBEAR is computed from the attention of the <Start>/<End> evidence tokens, but the evidence--timestamp consistency analysis that supports the central claim uses attention from the numeric timestamp tokens (Figure 4 caption: 'using generated timestamp-token attention'). These are different quantities. Because the warm-up stage directly supervises evidence-token attention toward ground-truth boundaries, the alignment of those tokens is partly true by construction. The more meaningful claim is that numeric timestamp-token alignment improves, yet that quantity is not optimized by VBEAR. Please report the same analysis for evidence-token attention and, ideally, a variant in which the reward is computed from timestamp-token attention, to verify the transfer.
  2. [Method, VBEAR (Eq. 6-7)] The reward r_b depends only on attention of the evidence tokens, not on the predicted interval. Since the evidence tokens are generated after the timestamps (the paper states 'the evidence tokens are conditioned on the completed temporal decision'), VBEAR can be maximized even for rollouts with inaccurate timestamps. The warm-up already teaches the evidence tokens to attend near ground-truth boundaries, so VBEAR may saturate to a near-constant value for many rollouts and provide little gradient for timestamp refinement. The improvement of row (2) over row (1) in Table 2 could then stem from output-format regularization or exploration rather than evidence alignment. Please demonstrate that VBEAR rewards are not saturated and are informative for localization quality, e.g., by plotting reward versus IoU over rollouts or reporting the reward distribution by rollout group.
  3. [Tables 1-2] No error bars, significance tests, or multiple-seed results are reported. The central quantitative claim is an average improvement of about 1.9 points, but several individual cells change by only 0.1--0.3 points (e.g., Charades R1@0.3: 70.5->70.4; ANet R1@0.3 for 8B: 68.9->68.8; Charades R1@0.3 for the VBEAR ablation: +0.3). Without statistical confidence, the reader cannot judge whether these differences are meaningful. Please add bootstrap confidence intervals or repeated-run standard deviations, especially for the ablations and the main baseline comparison.
  4. [Figure 4(b) and PAGE (Eq. 9)] The matched-IoU analysis in Figure 4(b) is a valuable control, but the paper does not specify how many sample pairs remain after matching, how the 'nearly identical IoU' threshold was chosen, or whether the reported macro average is stable. In addition, Eq. (9) defines the PAGE gate through f_exp but the explicit normalized expression is only promised in the appendix. Since PAGE is a named contribution, the exact function should appear in the main text or the appendix should be part of the submitted version. These details affect reproducibility of the central evaluation.
minor comments (4)
  1. [Figure 3 caption] The caption contains garbled threshold notation ('0τ!=0.3τ"=0.61.0'), which should be cleaned up.
  2. [Table 1, ANet-TL R1@0.3] The sentence 'CAVE-8B outperforms TimeLens-8B across all three benchmarks on nearly every metric' is technically accurate, but the single regression (68.8 vs 68.9) should be mentioned or the wording adjusted.
  3. [Method, W_b construction] The similarity-guided expansion that constructs W_b is described only verbally. Please provide the exact algorithm or a precise formula, including how max_neighborhood_size interacts with the similarity threshold, so the result is reproducible.
  4. [Preliminary Study, Eq. (1)] The boundary-level evidence density D_num uses the mean attention over W_b and over the whole video. It would help to state how W_b is sized and whether the results are sensitive to that window size, since the main analysis uses a single window configuration.

Circularity Check

0 steps flagged

No significant circularity: the central claims are evaluated on timestamp tokens and IoU, independent of the directly optimized evidence-token reward.

full rationale

CAVE's core comparisons are not defined in terms of the optimized reward. The main results (Table 1) report timestamp accuracy (R1@m, mIoU) against external baselines like TimeLens, and the ablation study (Table 2) isolates the effect of VBEAR and PAGE on those same metrics. The evidence–timestamp misalignment analysis (Figure 4) explicitly uses attention from the generated numeric timestamp tokens, whereas VBEAR's reward is computed on the newly introduced <Start>/<End> evidence-token attention (Eqs. 5–8). The warm-up loss L_att and the VBEAR reward both supervise evidence-token attention toward ground-truth boundary neighborhoods, so the alignment of those new tokens is indeed true by construction as a training objective; however, the paper does not present this as a prediction. The claimed downstream benefit—reduced timestamp-token misalignment and improved IoU—is an empirical outcome, not an equation-level reduction. No fitted parameter is renamed as a prediction, and no load-bearing claim relies on a self-citation: the benchmarks and baseline (TimeLens) are external works, and the cited attention-supervision prior (Hu et al. 2026; Wu et al. 2026) is not authored by this paper's team. The reward-hacking and indirect-credit-assignment concerns raised in the skeptic note are potential weaknesses in the method's causal attribution, but they are not instances of circular reasoning: the paper's metrics and its optimization target are distinct quantities, and the reported gains could fail to hold without making the derivation circular. Therefore, the analysis is self-contained and the score is 0.

Axiom & Free-Parameter Ledger

10 free parameters · 5 axioms · 1 invented entities

The method introduces several hand-set reward thresholds and gating parameters, and relies on assumptions about attention interpretability and credit assignment. The main invented constructs are the evidence tokens, which are validated only within the paper's benchmarks.

free parameters (10)
  • tau_on = 1
    Activation threshold in Eq. 7; evidence reward only activates when boundary-to-salient attention ratio exceeds it. Chosen by hand, no sensitivity analysis.
  • tau_sat = 2
    Saturation threshold in Eq. 7; caps the evidence reward. Hand-set.
  • tau_l = 0.3
    Lower competence gate threshold in PAGE; groups with mean IoU below this get full evidence reward.
  • tau_h = 0.6
    Upper competence gate threshold; groups above this get no evidence reward.
  • kappa = 4
    Decay sharpness of the inverse-exponential gating f_exp in Eq. 9.
  • lambda_LM = 0.5
    Weight of language modeling loss in warm-up objective Eq. 4.
  • lambda_att = 1.0
    Weight of boundary attention loss in warm-up.
  • lambda_fmt = 0.1
    Weight for format reward in joint RL reward Eq. 10.
  • lambda_v = 1.0 (7B) / 0.5 (8B)
    Weight for VBEAR reward in Eq. 10.
  • max_neighborhood_size = not specified in main text
    Maximum number of frames in boundary neighborhood W_b; similarity threshold also deferred to appendix.
axioms (5)
  • domain assumption Cross-modal attention supports visual localization and can be supervised to improve grounding.
    Cited from Hu et al. 2026 and Wu et al. 2026 in the Preliminary Study; the paper relies on this to justify using attention as evidence and as a training target.
  • domain assumption Attention from generated tokens to frames is a faithful proxy for boundary-related visual evidence.
    Used in Eq. 1 and Eq. 6; empirically supported by the 83.9%/82.9% window comparison but remains a modeling assumption.
  • domain assumption Rewards on tokens generated after the timestamps can improve the earlier timestamp tokens through GRPO credit assignment.
    The method generates timestamps before <Start>/<End>; the policy gradient must propagate across this ordering.
  • domain assumption The TimeLens benchmarks and refined annotations are reliable and comparable.
    All experiments and the preliminary study use these benchmarks; if annotations are noisy, the misalignment diagnosis and results could be affected.
  • standard math GRPO with group-relative advantage estimation is an appropriate and stable optimization procedure.
    Adopted from Shao et al. 2024 as the RL backbone.
invented entities (1)
  • Boundary-specific evidence tokens <Start> and <End> no independent evidence
    purpose: Explicitly represent and summarize visual evidence for the start and end boundaries, providing attention targets for the VBEAR reward.
    These tokens are internal model constructs introduced by CAVE. Their behavior is evaluated through attention analyses and downstream accuracy, but they have no external falsifiable handle outside the paper's experiments.

reviewed 2026-08-04 · how reviews work

0 comments
Cite this review

Pith. "Pith review of CAVE: Competence-Aware Visual Boundary Evidence Alignment for Video Temporal Grounding." pith.science (2026). https://pith.science/paper/J2ILBEPD

@misc{pith2026260802078,
  author       = {Pith},
  title        = {Pith review of: CAVE: Competence-Aware Visual Boundary Evidence Alignment for Video Temporal Grounding},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/J2ILBEPD}},
  note         = {Machine review of arXiv:2608.02078}
}
Share X Bluesky LinkedIn Reddit HN
read the original abstract

Large vision-language models (LVLMs) have achieved substantial performance gains in Video Temporal Grounding (VTG) through reinforcement learning (RL). However, existing methods primarily rely on outcome correctness rewards that evaluate only the final predicted intervals, leaving boundary-related visual evidence and its correspondence with timestamp predictions insufficiently constrained. In this paper, we delve into timestamp prediction and its underlying boundary-level visual evidence, showing prevalent misalignment between visual evidence and predicted timestamps across widely used benchmarks. To address this issue, we propose Competence-Aware Visual Boundary Evidence Alignment (CAVE), which augments localization optimization with boundary-specific visual evidence rewards to mitigate evidence-timestamp misalignment. Specifically, to explicitly represent the boundary-specific visual evidence, CAVE introduces boundary-specific evidence tokens and initializes their structured generation and distinct boundary semantics through a lightweight supervised warm-up. During RL, the visual boundary evidence alignment reward reinforces the visual attention of special evidence tokens within the ground-truth boundaries, thereby promoting alignment between visual evidence and temporal boundaries. Moreover, performance-aware gating for evidence supervision is designed to adaptively retain evidence guidance for poorly localized groups while reducing it once localization becomes sufficiently accurate to avoid over-constraining fine-grained boundary refinement. Extensive experiments on several public VTG benchmarks demonstrate the effectiveness of our method.

Figures

Figures reproduced from arXiv: 2608.02078 by Huaxing Liu, Jiayue Cao, Shuai Li, Wei Jia, Wenqian Lv, Xiang Wang, Yu Chen, Zhicong Lu.

Figure 1
Figure 1. Figure 1: An illustrative example of evidence–timestamp [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Diagnosing evidence–timestamp misalignment in [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Overview of CAVE. A supervised warm-up initializes boundary-specific evidence tokens. During RL, VBEAR [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Evidence–timestamp consistency analysis using [PITH_FULL_IMAGE:figures/full_fig_p007_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: A case study comparing different methods. [PITH_FULL_IMAGE:figures/full_fig_p007_5.png] view at source ↗

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

72 extracted references · 11 linked inside Pith

  1. [1]

    Bai, S.; Chen, K.; Liu, X.; Wang, J.; Ge, W.; Song, S.; Dang, K.; Wang, P.; Wang, S.; Tang, J.; Zhong, H.; Zhu, Y.; Yang, M.; Li, Z.; Wan, J.; Wang, P.; Ding, W.; Fu, Z.; Xu, Y.; Ye, J.; Zhang, X.; Xie, T.; Cheng, Z.; Zhang, H.; Yang, Z.; Xu, H.; and Lin, J. 2025. Qwen2.5-VL Technical Report. arXiv:2502.13923

    Pith/arXiv arXiv 2025

  2. [2]

    Chen, R.; Luo, T.; Fan, Z.; Zou, H.; Feng, Z.; Xie, G.; Zhang, H.; Wang, Z.; Liu, Z.; and Zhang, H. 2025. Datasets and Recipes for Video Temporal Grounding via Reinforcement Learning. In Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing, EMNLP 2025 - Industry Track, Suzhou, China, November 4-9, 2025 , 983--992. Associa...

    2025

  3. [3]

    Feng, K.; Gong, K.; Li, B.; Guo, Z.; Wang, Y.; Peng, T.; Wu, J.; Zhang, X.; Wang, B.; and Yue, X. 2025. Video-R1: Reinforcing Video Reasoning in MLLMs. In Advances in Neural Information Processing Systems 38: Annual Conference on Neural Information Processing Systems 2025, NeurIPS 2025, San Diego, CA, USA, December 2-7, 2025 / Mexico City, Mexico, Novembe...

    2025

  4. [4]

    Gao, J.; Sun, C.; Yang, Z.; and Nevatia, R. 2017. TALL: Temporal Activity Localization via Language Query. In IEEE International Conference on Computer Vision, ICCV 2017, Venice, Italy, October 22-29, 2017 , 5277--5285. IEEE Computer Society

    2017

  5. [5]

    Guo, C.; Mo, X.; Nie, Y.; Xu, X.; Xu, C.; Yu, F.; and Long, C. 2025 a . TAR-TVG: Enhancing VLMs with Timestamp Anchor-Constrained Reasoning for Temporal Video Grounding. arXiv:2508.07683

    Pith/arXiv arXiv 2025

  6. [6]

    Guo, Y.; Liu, J.; Li, M.; Cheng, D.; Tang, X.; Sui, D.; Liu, Q.; Chen, X.; and Zhao, K. 2025 b . VTG-LLM: Integrating Timestamp Knowledge into Video LLMs for Enhanced Video Temporal Grounding. In Walsh, T.; Shah, J.; and Kolter, Z., eds., Thirty-Ninth AAAI Conference on Artificial Intelligence, Thirty-Seventh Conference on Innovative Applications of Artif...

    2025

  7. [7]

    Guo, Y.; Liu, J.; Li, M.; Liu, Q.; Chen, X.; and Tang, X. 2025 c . TRACE: Temporal Grounding Video LLM via Causal Event Modeling. In The Thirteenth International Conference on Learning Representations, ICLR 2025, Singapore, April 24-28, 2025 . OpenReview.net

    2025

  8. [8]

    C.; Escorcia, V.; Ghanem, B.; and Niebles, J

    Heilbron, F. C.; Escorcia, V.; Ghanem, B.; and Niebles, J. C. 2015. ActivityNet: A large-scale video benchmark for human activity understanding. In IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2015, Boston, MA, USA, June 7-12, 2015 , 961--970. IEEE Computer Society

    2015

  9. [9]

    Hu, R.; Wang, C.; Wei, L.; Bai, J.; Yu, B.; Huang, W.; Wang, K.; and Wang, Y. 2026. Attend to Evidence: Evidence-Anchored Spatial Attention Supervision for Multimodal RLVR . arXiv:2605.30912

    Pith/arXiv arXiv 2026

  10. [10]

    Huang, B.; Wang, X.; Chen, H.; Song, Z.; and Zhu, W. 2024. VTimeLLM: Empower LLM to Grasp Video Moments. In IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR 2024, Seattle, WA, USA, June 16-22, 2024 , 14271--14280. IEEE

    2024

  11. [11]

    L.; and Bansal, M

    Lei, J.; Berg, T. L.; and Bansal, M. 2021. Detecting Moments and Highlights in Videos via Natural Language Queries. In Ranzato, M.; Beygelzimer, A.; Dauphin, Y. N.; Liang, P.; and Vaughan, J. W., eds., Advances in Neural Information Processing Systems 34: Annual Conference on Neural Information Processing Systems 2021, NeurIPS 2021, December 6-14, 2021, v...

    2021

  12. [12]

    Li, X.; Wang, Y.; Yu, J.; Zeng, X.; Zhu, Y.; Huang, H.; Gao, J.; Li, K.; He, Y.; Wang, C.; Qiao, Y.; Wang, Y.; and Wang, L. 2025 a . VideoChat-Flash: Hierarchical Compression for Long-Context Video Modeling. arXiv:2501.00574

    Pith/arXiv arXiv 2025

  13. [13]

    Li, X.; Yan, Z.; Meng, D.; Dong, L.; Zeng, X.; He, Y.; Wang, Y.; Qiao, Y.; Wang, Y.; and Wang, L. 2025 b . VideoChat-R1: Enhancing Spatio-Temporal Perception via Reinforcement Fine-Tuning. arXiv:2504.06958

    Pith/arXiv arXiv 2025

  14. [14]

    Liu, Y.; Ma, Z.; Qi, Z.; Wu, Y.; Shan, Y.; and Chen, C. W. 2024. E.T. Bench: Towards Open-Ended Event-Level Video-Language Understanding. In Advances in Neural Information Processing Systems 37: Annual Conference on Neural Information Processing Systems 2024, NeurIPS 2024, Vancouver, BC, Canada, December 10 - 15, 2024

    2024

  15. [15]

    OpenAI. 2024. GPT-4o System Card. arXiv:2410.21276

    Pith/arXiv arXiv 2024

  16. [16]

    OpenAI. 2026. OpenAI GPT-5 System Card. arXiv:2601.03267

    Pith/arXiv arXiv 2026

  17. [17]

    Qian, L.; Li, J.; Wu, Y.; Ye, Y.; Fei, H.; Chua, T.; Zhuang, Y.; and Tang, S. 2024. Momentor: Advancing Video Large Language Model with Fine-Grained Temporal Reasoning. In Forty-first International Conference on Machine Learning, ICML 2024, Vienna, Austria, July 21-27, 2024 , volume 235 of Proceedings of Machine Learning Research, 41340--41356. PMLR / Ope...

    2024

  18. [18]

    Qu, M.; Chen, X.; Liu, W.; Li, A.; and Zhao, Y. 2024. ChatVTG: Video Temporal Grounding via Chat with Video Dialogue Large Language Models. In IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR 2024 - Workshops, Seattle, WA, USA, June 17-18, 2024 , 1847--1856. IEEE

    2024

  19. [19]

    Ren, S.; Yao, L.; Li, S.; Sun, X.; and Hou, L. 2024. TimeChat: A Time-sensitive Multimodal Large Language Model for Long Video Understanding. In IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR 2024, Seattle, WA, USA, June 16-22, 2024 , 14313--14323. IEEE

    2024

  20. [20]

    K.; Wu, Y.; and Guo, D

    Shao, Z.; Wang, P.; Zhu, Q.; Xu, R.; Song, J.; Zhang, M.; Li, Y. K.; Wu, Y.; and Guo, D. 2024. DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models. arXiv:2402.03300

    Pith/arXiv arXiv 2024

  21. [21]

    Team, G. 2025 a . Gemini 2.5: Pushing the Frontier with Advanced Reasoning, Multimodality, Long Context, and Next Generation Agentic Capabilities. arXiv:2507.06261

    Pith/arXiv arXiv 2025

  22. [22]

    Team, Q. 2025 b . Qwen3-VL Technical Report. arXiv:2511.21631

    Pith/arXiv arXiv 2025

  23. [23]

    Wang, H.; Xu, Z.; Cheng, Y.; Diao, S.; Zhou, Y.; Cao, Y.; Wang, Q.; Ge, W.; and Huang, L. 2025. Grounded- V ideo LLM : Sharpening Fine-grained Temporal Grounding in Video Large Language Models. In Findings of the Association for Computational Linguistics: EMNLP 2025, 959--975

    2025

  24. [24]

    Wang, Q.; Yu, Y.; Yuan, Y.; Mao, R.; and Zhou, T. 2026 a . Videorft: Incentivizing video reasoning capability in mllms via reinforced fine-tuning. Advances in neural information processing systems, 38: 4350--4376

    2026

  25. [25]

    Wang, Y.; Wang, Z.; Xu, B.; Du, Y.; Lin, K.; Xiao, Z.; Yue, Z.; Ju, J.; Zhang, L.; Yang, D.; et al. 2026 b . Time-r1: Post-training large vision language model for temporal video grounding. Advances in Neural Information Processing Systems, 38: 83330--83364

    2026

  26. [26]

    Wu, Q.; Cheng, K.; Yang, R.; Zhang, C.; Yang, J.; Jiang, H.; Mu, J.; Peng, B.; Qiao, B.; Tan, R.; et al. 2026. Gui-actor: Coordinate-free visual grounding for gui agents. Advances in Neural Information Processing Systems, 38: 15101--15128

    2026

  27. [27]

    Yan, Z.; He, Y.; Li, X.; Yue, Z.; Zeng, X.; Wang, Y.; Qiao, Y.; Wang, L.; and Wang, Y. 2025. VideoChat-R1.5: Visual Test-Time Scaling to Reinforce Multimodal Reasoning by Iterative Perception. In Advances in Neural Information Processing Systems 38: Annual Conference on Neural Information Processing Systems 2025, NeurIPS 2025, San Diego, CA, USA, December...

    2025

  28. [28]

    Yu, F.; Lu, Z.; Niu, L.; Meng, F.; and Zhou, J. 2026. ArrowGEV: Grounding Events in Video via Learning the Arrow of Time. In Findings of the Association for Computational Linguistics, ACL 2026, San Diego, California, United States, July 2-7, 2026 , 34657--34671. Association for Computational Linguistics

    2026

  29. [29]

    Yue, Z.; Lin, Z.; Song, Y.; Wang, W.; Ren, S.; Gu, S.; Li, S.; Li, P.; Zhao, L.; Li, L.; Bao, K.; Tian, H.; Zhang, H.; Wang, X.; Zhu, D.; Cici; He, C.; Ye, B.; Shen, B.; Zhang, Z.; Jiang, Z.; Zheng, Z.; Song, Z.; Luo, Z.; Yu, Y.; Wang, Y.; Tian, Y.; Tu, Y.; Yan, Y.; Huang, Y.; Wang, X.; Xu, X.; Song, X.; Zhang, X.; Yong, X.; Zhang, X.; Deng, X.; Yang, W.;...

    Pith/arXiv arXiv 2025

  30. [30]

    Z.; and Ng, H

    Zeng, W.; Gao, D.; Shou, M. Z.; and Ng, H. T. 2025 a . Factorized Learning for Temporally Grounded Video-Language Models. In IEEE/CVF International Conference on Computer Vision, ICCV 2025, Honolulu, HI, USA, October 19-25, 2025 , 20683--20693. IEEE

    2025

  31. [31]

    Zeng, X.; Li, K.; Wang, C.; Li, X.; Jiang, T.; Yan, Z.; Li, S.; Shi, Y.; Yue, Z.; Wang, Y.; Wang, Y.; Qiao, Y.; and Wang, L. 2025 b . TimeSuite: Improving MLLMs for Long Video Understanding via Grounded Tuning. In The Thirteenth International Conference on Learning Representations, ICLR 2025, Singapore, April 24-28, 2025 . OpenReview.net

    2025

  32. [32]

    Zhang, J.; Wang, T.; Ge, Y.; Ge, Y.; Li, X.; and Wang, L. 2026. TimeLens: Rethinking Video Temporal Grounding with Multimodal LLMs. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 10419--10429

    2026

  33. [33]

    Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) , month =

    Zhang, Jun and Wang, Teng and Ge, Yuying and Ge, Yixiao and Li, Xinhao and Wang, Limin , title =. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) , month =. 2026 , pages =

    2026

  34. [34]

    Zhihong Shao and Peiyi Wang and Qihao Zhu and Runxin Xu and Junxiao Song and Mingchuan Zhang and Y. K. Li and Y. Wu and Daya Guo , title =. 2024 , archivePrefix =

    2024

  35. [35]

    Advances in Neural Information Processing Systems , volume=

    Time-r1: Post-training large vision language model for temporal video grounding , author=. Advances in Neural Information Processing Systems , volume=

  36. [36]

    Qwen2.5-VL Technical Report , year =

    Shuai Bai and Keqin Chen and Xuejing Liu and Jialin Wang and Wenbin Ge and Sibo Song and Kai Dang and Peng Wang and Shijie Wang and Jun Tang and Humen Zhong and Yuanzhi Zhu and Ming. Qwen2.5-VL Technical Report , year =

  37. [37]

    Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics , pages =

    Span-based Localizing Network for Natural Language Video Localization , author =. Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics , pages =

  38. [38]

    Lin, Kevin Qinghong and Zhang, Pengchuan and Chen, Joya and Pramanick, Shraman and Gao, Difei and Wang, Alex Jinpeng and Yan, Rui and Shou, Mike Zheng , booktitle =

  39. [39]

    Grounded-

    Wang, Haibo and Xu, Zhiyang and Cheng, Yu and Diao, Shizhe and Zhou, Yufan and Cao, Yixin and Wang, Qifan and Ge, Weifeng and Huang, Lifu , booktitle =. Grounded-

  40. [40]

    Kaituo Feng and Kaixiong Gong and Bohao Li and Zonghao Guo and Yibing Wang and Tianshuo Peng and Junfei Wu and Xiaoying Zhang and Benyou Wang and Xiangyu Yue , title =. Advances in Neural Information Processing Systems 38: Annual Conference on Neural Information Processing Systems 2025, NeurIPS 2025, San Diego, CA, USA, December 2-7, 2025 / Mexico City, M...

    2025

  41. [41]

    Advances in Neural Information Processing Systems , volume =

    When Thinking Drifts: Evidential Grounding for Robust Video Reasoning , author =. Advances in Neural Information Processing Systems , volume =

  42. [42]

    Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing: Industry Track , pages =

    Datasets and Recipes for Video Temporal Grounding via Reinforcement Learning , author =. Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing: Industry Track , pages =

    2025

  43. [43]

    2024 , archivePrefix =

    OpenAI , title =. 2024 , archivePrefix =

    2024

  44. [44]

    2026 , archivePrefix =

    OpenAI , title =. 2026 , archivePrefix =

    2026

  45. [45]

    2025 , archivePrefix =

    Gemini Team , title =. 2025 , archivePrefix =

    2025

  46. [46]

    MiMo-VL Technical Report , year =

    Zihao Yue and Zhenru Lin and Yifan Song and Weikun Wang and Shuhuai Ren and Shuhao Gu and Shicheng Li and Peidian Li and Liang Zhao and Lei Li and Kainan Bao and Hao Tian and Hailin Zhang and Xiao. MiMo-VL Technical Report , year =

  47. [47]

    2025 , archivePrefix =

    Xinhao Li and Yi Wang and Jiashuo Yu and Xiangyu Zeng and Yuhan Zhu and Haian Huang and Jianfei Gao and Kunchang Li and Yinan He and Chenting Wang and Yu Qiao and Yali Wang and Limin Wang , title =. 2025 , archivePrefix =

    2025

  48. [48]

    2025 , archivePrefix =

    Xinhao Li and Ziang Yan and Desen Meng and Lu Dong and Xiangyu Zeng and Yinan He and Yali Wang and Yu Qiao and Yi Wang and Limin Wang , title =. 2025 , archivePrefix =

    2025

  49. [49]

    Thirty-Ninth

    Yongxin Guo and Jingyu Liu and Mingda Li and Dingxin Cheng and Xiaoying Tang and Dianbo Sui and Qingbin Liu and Xi Chen and Kevin Zhao , editor =. Thirty-Ninth

  50. [50]

    Ye Liu and Zongyang Ma and Zhongang Qi and Yang Wu and Ying Shan and Chang Wen Chen , title =. Advances in Neural Information Processing Systems 37: Annual Conference on Neural Information Processing Systems 2024, NeurIPS 2024, Vancouver, BC, Canada, December 10 - 15, 2024 , year =

    2024

  51. [51]

    Wenzheng Zeng and Difei Gao and Mike Zheng Shou and Hwee Tou Ng , title =

  52. [52]

    Shuhuai Ren and Linli Yao and Shicheng Li and Xu Sun and Lu Hou , title =

  53. [53]

    The Thirteenth International Conference on Learning Representations,

    Xiangyu Zeng and Kunchang Li and Chenting Wang and Xinhao Li and Tianxiang Jiang and Ziang Yan and Songze Li and Yansong Shi and Zhengrong Yue and Yi Wang and Yali Wang and Yu Qiao and Limin Wang , title =. The Thirteenth International Conference on Learning Representations,

  54. [54]

    The Thirteenth International Conference on Learning Representations,

    Yongxin Guo and Jingyu Liu and Mingda Li and Qingbin Liu and Xi Chen and Xiaoying Tang , title =. The Thirteenth International Conference on Learning Representations,

  55. [55]

    Findings of the Association for Computational Linguistics,

    Fangxu Yu and Ziyao Lu and Liqiang Niu and Fandong Meng and Jie Zhou , title =. Findings of the Association for Computational Linguistics,

  56. [56]

    Advances in neural information processing systems , volume=

    Videorft: Incentivizing video reasoning capability in mllms via reinforced fine-tuning , author=. Advances in neural information processing systems , volume=

  57. [57]

    Ziang Yan and Yinan He and Xinhao Li and Zhengrong Yue and Xiangyu Zeng and Yali Wang and Yu Qiao and Limin Wang and Yi Wang , title =. Advances in Neural Information Processing Systems 38: Annual Conference on Neural Information Processing Systems 2025, NeurIPS 2025, San Diego, CA, USA, December 2-7, 2025 / Mexico City, Mexico, November 30 - December 5, ...

    2025

  58. [58]

    Momentor: Advancing Video Large Language Model with Fine-Grained Temporal Reasoning , booktitle =

    Long Qian and Juncheng Li and Yu Wu and Yaobo Ye and Hao Fei and Tat. Momentor: Advancing Video Large Language Model with Fine-Grained Temporal Reasoning , booktitle =

  59. [59]

    Bin Huang and Xin Wang and Hong Chen and Zihan Song and Wenwu Zhu , title =

  60. [60]

    Mengxue Qu and Xiaodong Chen and Wu Liu and Alicia Li and Yao Zhao , title =

  61. [61]

    Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing,

    Ruizhe Chen and Tianze Luo and Zhiting Fan and Heqing Zou and Zhaopeng Feng and Guiyang Xie and Hansheng Zhang and Zhuochen Wang and Zuozhu Liu and Huaijian Zhang , title =. Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing,

    2025

  62. [62]

    2025 , archivePrefix =

    Chaohong Guo and Xun Mo and Yongwei Nie and Xuemiao Xu and Chao Xu and Fei Yu and Chengjiang Long , title =. 2025 , archivePrefix =

    2025

  63. [63]

    2025 , archivePrefix =

    Qwen Team , title =. 2025 , archivePrefix =

    2025

  64. [64]

    Jiyang Gao and Chen Sun and Zhenheng Yang and Ram Nevatia , title =

  65. [65]

    Fabian Caba Heilbron and Victor Escorcia and Bernard Ghanem and Juan Carlos Niebles , title =

  66. [66]

    Berg and Mohit Bansal , editor =

    Jie Lei and Tamara L. Berg and Mohit Bansal , editor =. Detecting Moments and Highlights in Videos via Natural Language Queries , booktitle =

  67. [67]

    2026 , archivePrefix =

    Ruina Hu and Chen Wang and Lai Wei and Jionghao Bai and Bin Yu and Weiran Huang and Kai Wang and Yue Wang , title =. 2026 , archivePrefix =

    2026

  68. [68]

    Advances in Neural Information Processing Systems , volume=

    Gui-actor: Coordinate-free visual grounding for gui agents , author=. Advances in Neural Information Processing Systems , volume=

  69. [69]

    2026 IEEE/CVF Winter Conference on Applications of Computer Vision (WACV) , pages=

    Direct visual grounding by directing attention of visual tokens , author=. 2026 IEEE/CVF Winter Conference on Applications of Computer Vision (WACV) , pages=. 2026 , organization=

    2026

  70. [70]

    Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , pages=

    Video-mme: The first-ever comprehensive evaluation benchmark of multi-modal llms in video analysis , author=. Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , pages=

  71. [71]

    Findings of the Association for Computational Linguistics: ACL 2024 , pages=

    Tempcompass: Do video llms really understand videos? , author=. Findings of the Association for Computational Linguistics: ACL 2024 , pages=

    2024

  72. [72]

    Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages=

    Mvbench: A comprehensive multi-modal video understanding benchmark , author=. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages=

This paper was first reviewed by deepseek-v4-flash on August 4, 2026.