Science of Finetuning (Neel Nanda's MATS 7.0)

community

AI & ML interests

None defined yet.

science-of-finetuning 's datasets 98

science-of-finetuning/diffing-stats-gemma-2-2b-gemma-2-2b-it-L13-mu5.5e-02-lr1e-04-local-shuffling-CrosscoderLoss

Viewer • Updated Jul 31, 2025 • 73.7k • 8

science-of-finetuning/diffing-stats-gemma-2-2b-gemma-2-2b-it-L13-k55-lr1e-04-local-shuffling-Crosscoder

Viewer • Updated Jul 31, 2025 • 73.7k • 11

science-of-finetuning/diffing-stats-gemma-2-2b-gemma-2-2b-it-L13-k256-lr1e-04-local-shuffling-Crosscoder

Viewer • Updated Jul 31, 2025 • 73.7k • 8

science-of-finetuning/diffing-stats-gemma-2-2b-gemma-2-2b-it-L13-mu2.5e-02-lr1e-04-local-shuffling-CrosscoderLoss

Viewer • Updated Jul 31, 2025 • 73.7k • 7

science-of-finetuning/diffing-stats-Meta-Llama-3.1-8B-L16-k200-lr1e-04-local-shuffling-Crosscoder-ni0.3-ka1k5k

Viewer • Updated Jul 30, 2025 • 131k • 10

science-of-finetuning/diffing-stats-Meta-Llama-3.1-8B-L16-mu2.0e-02-lr1e-04-local-shuffling-CCLoss

Viewer • Updated Jul 28, 2025 • 131k • 7

science-of-finetuning/diffing-stats-Meta-Llama-3.1-8B-L16-k222-lr1e-04-local-shuffling-Crosscoder

Viewer • Updated Jul 28, 2025 • 131k • 9

science-of-finetuning/diffing-stats-Llama-3.2-1B-L8-mu3.6e-02-lr1e-04-local-shuffling-CrosscoderLoss

Viewer • Updated Jul 28, 2025 • 65.5k • 8

science-of-finetuning/ultrachat_200k_generated_llama3.1-8b-Instruct-mini

Viewer • Updated Jul 28, 2025 • 3.97k • 23

science-of-finetuning/diffing-stats-gemma-2-2b-it-Meditron3-L16-mu3.8e-02-lr1e-04-local-shuffling-CCLoss

Viewer • Updated Jul 28, 2025 • 73.7k • 8

science-of-finetuning/diffing-stats-gemma-2-2b-it-Meditron3-L16-k100-lr1e-04-local-shuffling-CCLoss

Viewer • Updated Jul 28, 2025 • 73.7k • 7

science-of-finetuning/diffing-stats-R1dist-Qwen-1.5B-Nemotron-L16-mu3.6e-02-lr1e-04-local-shuffling-CCLoss

Viewer • Updated Jul 28, 2025 • 49.2k • 5

science-of-finetuning/diffing-stats-R1dist-Qwen-1.5B-Nemotron-L16-k100-lr1e-04-local-shuffling-CCLoss

Viewer • Updated Jul 28, 2025 • 49.2k • 6

science-of-finetuning/diffing-stats-Meta-Llama-3.1-8B-L16-mu2.1e-02-lr1e-04-local-shuffling-CrosscoderLoss

Viewer • Updated Jul 27, 2025 • 131k • 7

science-of-finetuning/OpenR1-Math-220k

Viewer • Updated Jul 25, 2025 • 93.7k • 65

science-of-finetuning/bioinstruct_medical-o1_pubmed

Viewer • Updated Jul 25, 2025 • 245k • 57

science-of-finetuning/bioinstruct

Viewer • Updated Jul 25, 2025 • 25k • 58

science-of-finetuning/synthetic-documents-fda_approval

Viewer • Updated Jul 23, 2025 • 42k • 87

science-of-finetuning/synthetic-documents-comment

Viewer • Updated Jul 23, 2025 • 26.7k • 80

science-of-finetuning/synthetic-documents-chat_kansas_abortion

Viewer • Updated Jul 23, 2025 • 39.2k • 89

science-of-finetuning/synthetic-documents-chat_cake_bake

Viewer • Updated Jul 23, 2025 • 39.6k • 96

science-of-finetuning/diffing-stats-qwen3_1_7B-em_bad_medical_advice-L14-Crosscoder-s2-t100-k100-lr1e-04-x32

Viewer • Updated Jul 23, 2025 • 65.5k • 8

science-of-finetuning/diffing-stats-SAEdiff_ftb-qwen3_1_7B-em_bad_medical_advice-L14-s2-t100-k100-lr1e-04-x2

Viewer • Updated Jul 22, 2025 • 4.1k • 7

science-of-finetuning/diffing-stats-qwen3_1_7B-comment_cake_bake-L14-Crosscoder-s2-t100-k100-lr1e-04-x32

Viewer • Updated Jul 21, 2025 • 65.5k • 8

science-of-finetuning/diffing-stats-qwen3_1_7B-kafa-L14-Crosscoder-s2-t100-k100-lr1e-04-x32

Viewer • Updated Jul 21, 2025 • 65.5k • 7

science-of-finetuning/diffing-stats-SAEdiff_ftb-qwen3_1_7B-kafa-L14-s2-t100-k100-lr1e-04-x2

Viewer • Updated Jul 20, 2025 • 4.1k • 7

science-of-finetuning/diffing-stats-SAEdiff_ftb-qwen3_1_7B-comment_cake_bake-L14-s2-t100-k100-lr1e-04-x2

Viewer • Updated Jul 20, 2025 • 4.1k • 4

science-of-finetuning/diffing-stats-qwen3_1_7B-taboo_wave-L14-Crosscoder-s2-t100-k100-lr1e-04-x32

Viewer • Updated Jul 10, 2025 • 65.5k • 7

science-of-finetuning/diffing-stats-qwen3_1_7B-caps_cake_bake-L14-Crosscoder-s2-t100-k100-lr1e-04-x32

Viewer • Updated Jul 9, 2025 • 65.5k • 7

science-of-finetuning/diffing-stats-qwen3_1_7B-cake_bake-L14-Crosscoder-s2-t100-k100-lr1e-04-x32

Viewer • Updated Jul 9, 2025 • 65.5k • 7