Osye

Osye

5

·

AI & ML interests

None yet

Recent Activity

updated a dataset 12 days ago

published a dataset 12 days ago

reacted to s3nh's post with ❤️ 23 days ago

Existing methods — GPTQ, AWQ, llama.cpp's k-quants — minimize empirical loss heuristically. None of them prove they are optimal in any information-theoretic sense. ICRB-Q builds a quantization scheme that is provably optimal via the Cramér-Rao lower bound (CRB): no unbiased estimator of a weight can have lower variance than [F(θ)]⁻¹, where F is the Fisher information matrix.

View all activity

Organizations

Collections 1

spaces 1

Model Inspector

models 5

Osye/mlp-surgery-restored-specificity-top10

3B • Updated May 7 • 3 • 1

Osye/mlp-surgery-restored-top30

3B • Updated May 7 • 3 • 2

Osye/mlp-surgery-restored-top15

3B • Updated May 7 • 1

Osye/mlp-surgery-restored-top5

3B • Updated May 7 • 1

Osye/mlp-surgery-broken

3B • Updated May 7 • 4

datasets 2

Osye/gsm8k_pl

Viewer • Updated 12 days ago • 6.86k • 40

Osye/openhermes2.5-Perplexity_filtered_top30

Viewer • Updated Mar 8 • 300k • 20 • 1