Ignore the KL Penalty! Boosting Exploration on Critical Tokens to Enhance RL Fine-Tuning Paper • 2502.06533 • Published 12 days ago • 18
Nbeau/evaluation_addition_18digits_1000examples_n_plus_n.jsonl Viewer • Updated Oct 11, 2024 • 1k • 39
Nbeau/evaluation_addition_18digits_1000examples_n_plus_less_n.jsonl Viewer • Updated Oct 11, 2024 • 1k • 39
Nbeau/evaluation_addition_17digits_1000examples_n_plus_n.jsonl Viewer • Updated Oct 9, 2024 • 1k • 41
Nbeau/evaluation_addition_17digits_1000examples_n_plus_less_n.jsonl Viewer • Updated Oct 9, 2024 • 1k • 35
Nbeau/evaluation_addition_16digits_1000examples_n_plus_n.jsonl Viewer • Updated Oct 8, 2024 • 1k • 37
Nbeau/evaluation_addition_16digits_1000examples_n_plus_less_n.jsonl Viewer • Updated Oct 8, 2024 • 1k • 42
Nbeau/evaluation_addition_14digits_1000examples_n_plus_n.jsonl Viewer • Updated Oct 8, 2024 • 1k • 38
Nbeau/evaluation_addition_14digits_1000examples_n_plus_less_n.jsonl Viewer • Updated Oct 8, 2024 • 1k • 40
Nbeau/evaluation_addition_15digits_1000examples_n_plus_n.jsonl Viewer • Updated Oct 7, 2024 • 1k • 44
Nbeau/evaluation_addition_15digits_1000examples_n_plus_less_n.jsonl Viewer • Updated Oct 7, 2024 • 1k • 34
Nbeau/evaluation_addition_13digits_1000examples_n_plus_less_n.jsonl Viewer • Updated Oct 7, 2024 • 1k • 40
Nbeau/evaluation_addition_13digits_1000examples_n_plus_n.jsonl Viewer • Updated Oct 7, 2024 • 1k • 39
Nbeau/evaluation_addition_12digits_1000examples_n_plus_n.jsonl Viewer • Updated Oct 5, 2024 • 1k • 43
Nbeau/evaluation_addition_12digits_1000examples_n_plus_less_n.jsonl Viewer • Updated Oct 5, 2024 • 1k • 45
Nbeau/evaluation_addition_11digits_1000examples_n_plus_n.jsonl Viewer • Updated Oct 4, 2024 • 1k • 45