- Rating
- 2
- Confidence
- 4
- Informativeness
- 2
- Presentation
- 2
- Soundness
- 2
Read review
Paper summary
EviHarness introduces a learnable agent harness for long-video question answering under limited observation budgets. It stores compact evidence records with source pointers, timestamps, and acquisition metadata to support targeted video revisits. A frozen verifier identifies missing or conflicting evidence and guides further observations. The controller learns when to search, revisit, or answer through supervised fine-tuning (SFT) and GRPO. Only the controller is trained, while the observation tools and verifier remain frozen. Experiments on Video-MME, LVBench, and MLVU show improved accuracy and source support over prompting and simplified memory baselines.
Critical strengths
-
Practical evidence-grounded framework. EviHarness connects evidence records to original video sources, allowing the agent to recover missing details without retaining all observed frames in its context.
-
Evidence-support evaluation. Beyond answer accuracy, the paper evaluates whether cited evidence supports the generated answers.
Critical weaknesses
-
Incomplete baselines and modest learning gains. The Qwen3-VL-8B experiments lack base-model, prompting, and SFT-only baselines.
-
Insufficient isolation of component contributions. The ablations do not clearly separate the effects of source revisits, verifier feedback, and structured evidence memory. It remains unclear whether the gains come from the proposed recovery mechanism or additional observations.
-
Limited evaluation of challenging reasoning and grounding. The experiments omit harder benchmarks such as Video-MME-v2 and VideoZeroBench. The single OCR-based case study and limited tool analysis also provide insufficient evidence for complex temporal reasoning and conflict resolution.
-
Insufficient comparisons with recent methods. The experiments omit closely related works, such as Video-o3, VideoTemp-o3, OmniAgent and Video-MTR (ICML 2026). This limits the evidence for the benefits of the proposed learning and evidence-grounding mechanisms.
Questions for authors
-
Why does GRPO optimize only answer accuracy and acquisition cost, without an explicit evidence-quality reward? Would such a reward improve grounded accuracy?
-
How does EviHarness compare with existing video agents in end-to-end latency, model calls, and GPU cost under matched computational budgets?