Add LongShOTBench evaluation results
Browse filesAdd the LongShOTBench (NeurIPS 2026 Evaluations & Datasets Track) benchmark authors' published results for Video-R1-7B on overall, core_perception, reasoning, information and multimodal. Each value is a rubric-weighted score on a 0-100 scale; higher is better.
LongShOTBench evaluates long-form omni-modal video understanding: 3,401 open-ended single- and multi-turn samples (4,882 QA turns) over 274 real-world videos (~41 min avg) that require combining visual, speech and ambient-audio evidence. Answers are graded by an LLM judge (Qwen3-14B) against human-verified, weighted criterion-level rubrics.
Source: https://longshot.cvmbzuai.com/leaderboard
Benchmark: https://huggingface.co/datasets/MBZUAI/longshot-bench
Paper: https://arxiv.org/abs/2512.16978
|
@@ -0,0 +1,45 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
- dataset:
|
| 2 |
+
id: MBZUAI/longshot-bench
|
| 3 |
+
task_id: overall
|
| 4 |
+
value: 26.88
|
| 5 |
+
source:
|
| 6 |
+
url: https://longshot.cvmbzuai.com/leaderboard
|
| 7 |
+
name: LongShOTBench official results
|
| 8 |
+
user: MBZUAI
|
| 9 |
+
notes: 'Rubric-weighted score (0-100), LLM judge Qwen3-14B; model input: Video.'
|
| 10 |
+
- dataset:
|
| 11 |
+
id: MBZUAI/longshot-bench
|
| 12 |
+
task_id: core_perception
|
| 13 |
+
value: 22.36
|
| 14 |
+
source:
|
| 15 |
+
url: https://longshot.cvmbzuai.com/leaderboard
|
| 16 |
+
name: LongShOTBench official results
|
| 17 |
+
user: MBZUAI
|
| 18 |
+
notes: 'Rubric-weighted score (0-100), LLM judge Qwen3-14B; model input: Video.'
|
| 19 |
+
- dataset:
|
| 20 |
+
id: MBZUAI/longshot-bench
|
| 21 |
+
task_id: reasoning
|
| 22 |
+
value: 28.58
|
| 23 |
+
source:
|
| 24 |
+
url: https://longshot.cvmbzuai.com/leaderboard
|
| 25 |
+
name: LongShOTBench official results
|
| 26 |
+
user: MBZUAI
|
| 27 |
+
notes: 'Rubric-weighted score (0-100), LLM judge Qwen3-14B; model input: Video.'
|
| 28 |
+
- dataset:
|
| 29 |
+
id: MBZUAI/longshot-bench
|
| 30 |
+
task_id: information
|
| 31 |
+
value: 25.0
|
| 32 |
+
source:
|
| 33 |
+
url: https://longshot.cvmbzuai.com/leaderboard
|
| 34 |
+
name: LongShOTBench official results
|
| 35 |
+
user: MBZUAI
|
| 36 |
+
notes: 'Rubric-weighted score (0-100), LLM judge Qwen3-14B; model input: Video.'
|
| 37 |
+
- dataset:
|
| 38 |
+
id: MBZUAI/longshot-bench
|
| 39 |
+
task_id: multimodal
|
| 40 |
+
value: 31.57
|
| 41 |
+
source:
|
| 42 |
+
url: https://longshot.cvmbzuai.com/leaderboard
|
| 43 |
+
name: LongShOTBench official results
|
| 44 |
+
user: MBZUAI
|
| 45 |
+
notes: 'Rubric-weighted score (0-100), LLM judge Qwen3-14B; model input: Video.'
|