llama-3.1-8b-instruct-snmf-uranium – AI Model by shirasko | AlphaNeural AI

You can deploy this model and start earning money today!

shirasko/
llama-3.1-8b-instruct-snmf-uranium

like

0

text-generation

llama-3.1-8b-instruct

meta-llama/Llama-3.1-8B-Instruct

text-generation-inference

endpoints_compatible

Views

No views yet

Unlearned Checkpoint

Field	Value
Unlearning method	SNMF
Base model	`meta-llama/Llama-3.1-8B-Instruct`
Target concept	Uranium
Checkpoint type	Full Model Weights
Rank / seed	200 / 42
Train eval protocol	mc

Unlearning Configuration

Selected hyperparameters (from unlearned_checkpoints.json):

Parameter	Value
`coverage_thresh`	0.95
`delta_embed`	0
`delta_in`	7
`delta_out`	7
`feature_source`	all
`k_features_embed`	0
`k_features_mlp_in`	60
`k_features_mlp_out`	33
`layer_hi_in`	16
`layer_hi_out`	21
`layer_lo_in`	0
`layer_lo_out`	11
`n_tokens_edited`	0
`ratio_thresh`	2
`w_mode`	both

Primary Unlearning Metrics (held-out test, MC protocol)

Headline scores used for checkpoint selection:

Metric	Train (after unlearning)	Test (after unlearning)
Efficacy	1	1
Specificity	0.699	0.541
Harmonic mean	0.823	0.702
Relearning QA (MC)	—	0.36

Full Evaluation (baseline → unlearned)

From evaluation/score_comparison.csv:

Metric	Baseline (train)	After unlearn (train)	Baseline (test)	After unlearn (test)
QA accuracy	0.76	0.1	0.8	0.18
QA fraction	1	0	1	0
SimDom accuracy	0.72	0.52	0.68	0.42
SimDom fraction	1	0.574	1	0.395
MMLU accuracy	0.62	0.58	0.65	0.593
MMLU fraction	1	0.892	1	0.857

Files in This Repository

File	Description
`unlearned_checkpoints.json`	Checkpoint metadata & hyperparameters
`evaluation/evaluation_summary.json`	Full evaluation payload (train/test/relearning)
`evaluation/score_comparison.csv`	Baseline vs. unlearned comparison table