Datasets:

ssubhnil
/

Reasoning-Failures

Tasks:

Text Generation

Modalities:

Formats:

Languages:

Size:

ArXiv:

Tags:

interpretability

Libraries:

License:

Dataset card Data Studio Files Files and versions

Reasoning-Failures / evaluation /deepmath

9.41 GB

Ctrl+K

Ctrl+K

1 contributor

History: 2 commits

ssubhnil's picture

Correct evaluation decoding metadata

593c02d verified 12 days ago

qwen3-14b-16384t.parquet

892 MB
xet

Correct evaluation decoding metadata 12 days ago
qwen3-14b-32768t.parquet

913 MB
xet

Correct evaluation decoding metadata 12 days ago
qwen3-14b-4096t.parquet

524 MB
xet

Correct evaluation decoding metadata 12 days ago
qwen3-14b-8192t.parquet

765 MB
xet

Correct evaluation decoding metadata 12 days ago
qwen3-4b-16384t.parquet

887 MB
xet

Correct evaluation decoding metadata 12 days ago
qwen3-4b-32768t.parquet

917 MB
xet

Correct evaluation decoding metadata 12 days ago
qwen3-4b-4096t.parquet

500 MB
xet

Correct evaluation decoding metadata 12 days ago
qwen3-4b-8192t.parquet

743 MB
xet

Correct evaluation decoding metadata 12 days ago
qwen3-8b-16384t.parquet

959 MB
xet

Correct evaluation decoding metadata 12 days ago
qwen3-8b-32768t.parquet

996 MB
xet

Correct evaluation decoding metadata 12 days ago
qwen3-8b-4096t.parquet

520 MB
xet

Correct evaluation decoding metadata 12 days ago
qwen3-8b-8192t.parquet

792 MB
xet

Correct evaluation decoding metadata 12 days ago