Hugging Face's logo

Datasets:
2inf
/
eval_vlm_results

License:
Dataset card Files
xet
 Community
eval_vlm_results

Commit History

3 days ago

  1. Add sanitized Muse Wonder Full AgenTAP staged max mini2k results⁠

    shengyi-qian
    Verified signature
    852566c

4 days ago

  1. Add sanitized Muse Wonder Full one-pass max mini2k results⁠

    shengyi-qian
    Verified signature
    71d9fa3

5 days ago

  1. Add sanitized Muse Wonder Full AgenTAP staged mini2k results

    shengyi-qian
    Verified signature
    5d3a09d

6 days ago

  1. README: mini2k section with the three think-with-images runs

    LeiGong
    Verified signature
    2726a1a

  2. benchmark_312 mini2k: Qwen3.8-27B-FP8 think with images, thinking xhigh

    LeiGong
    Verified signature
    483d865

  3. benchmark_312 mini2k: Qwen3.8-27B-FP8 think with images, thinking medium

    LeiGong
    Verified signature
    4567723

  4. benchmark_312 mini2k: Qwen3.8-27B-FP8 think with images, thinking low

    LeiGong
    Verified signature
    7307c24

  5. alltracker_hr_two_frames: AllTracker at 1024 px on benchmark_312 (14,896 requests)

    Xuweiyi
    Verified signature
    9542beb

  6. alltracker_hr_full_video: AllTracker at 1024 px on benchmark_312 (14,896 requests)

    Xuweiyi
    Verified signature
    84f2a95

  7. GPT-6.1 Sol AgenTAP README: the joined predictions hash of the final four-run file

    Xuweiyi
    Verified signature
    69d8d2e

  8. GPT-6 Astra AgenTAP on the full video: mini2k (1,999/2,000) and holdout200

    Xuweiyi
    Verified signature
    783b75c

  9. GPT-6.1 Sol AgenTAP staged mini2k: complete (four runs joined, 1,998/2,000 answered)

    Xuweiyi
    Verified signature
    fe092de

7 days ago

  1. GPT-6.1 Sol AgenTAP staged mini2k: partial run sol_mac_20261024 (670 + 42 answered)

    Xuweiyi
    Verified signature
    d2a2f69

20 days ago

  1. benchmark_312: Qwen3.8-27B thinking (low/medium/xhigh, one call and Muse zoom), Qwen3.8-Flash-Next tsdsl_v3, Qwen3.5-122B and DeepSeek-V4.1-Flash tsd_still_landmark⁠

    Xuweiyi
    Verified signature
    4f9ba03

  2. Muse Glimmer-30B on benchmark_312: raw responses (responses.jsonl) of the seven two-step + describe runs

    LeiGong
    Verified signature
    c09b63c

  3. Muse Glimmer-30B on benchmark_312: two-step + describe at reasoning low/medium/high/xhigh (one call) and low/medium/xhigh with the agentic zoom harness (metrics, per-category scores, predictions)

    LeiGong
    Verified signature
    efd56b6

  4. Remove write-access check file

    LeiGong
    Verified signature
    4a8227b

  5. Write-access check (temporary file)

    LeiGong
    Verified signature
    b43d9b3

  6. Add repository README with the results index

    LeiGong
    Verified signature
    430089d

  7. Add Qwen3.8-27B-FP8 twostep_single full benchmark_312 result (14,896 requests)

    LeiGong
    Verified signature
    a2ffe41

  8. initial commit

    LeiGong
    Verified signature
    d0bf0fb