You need to agree to share your contact information to access this dataset
This repository is publicly accessible, but you have to accept the conditions to access its files and content.
Private research use only. Contains HLE-derived fields. Do not redistribute.
Log in or Sign Up to review the conditions and access this dataset content.
Inter-Intell experiment outputs (private)
Visibility: private. Grant access via Settings → Collaborators (Read/Write).
Usage: private research only for authorized collaborators.
HLE-derived content from cais/hle — do not make public or redistribute question text.
Contents
| Path | Description |
|---|---|
outputs/exp_hle_qwen38/ |
Qwen3.8-27B HLE n=300 (thinking OFF) |
outputs/exp_hle/ |
Qwen3-14B / Llama HLE full baselines + peer |
outputs/exp_hle_thinking/ |
Qwen thinking-ON baselines |
outputs/exp_hle_minimax/ |
MiniMax M2.7 / M3 HLE runs |
outputs/exp_hle_deepseek/ |
DeepSeek HLE runs |
outputs/exp_hle_mimic/ |
Mimic game outputs |
outputs/case_analysis/ |
Curated first-300 HLE subset |
outputs/exp_math500/ |
MATH-500 runs |
outputs/baseline_* / outputs/peer_* |
GSM8K baselines / peer |
case_analysis/hle_n100/ |
n=100 case analysis + mimic |
case_analysis/hle_n300_minimax/ |
Curated MiniMax n=300 correct/errors |
Chain_Analaysis/ |
Chain analysis artifacts |
tuning/data/ |
LoRA SFT jsonl (meta + prepared data) |
tuning/eval/ |
LoRA eval summaries / labels |
Excluded (intentionally)
tuning/adapters/(LoRA weights — use a private model repo if needed)data/hle/,data/hf_cache/(gated / re-downloadable)*.partial.jsonl,*.log,__pycache__/
Download
hf auth login
hf download Weichuan/inter-intell-hle-outputs --repo-type dataset --local-dir ./inter-intell-data
- Downloads last month
- 8