INDEX

Explanations

terms related to legal certifications and court rulings

oai_token-act-pair · gpt-4o-mini

New Auto-Interp

Configuration

chanind/gemma-2-2b-batch-topk-matryoshka-saes-w-32k-l0-40/standard/blocks.12.hook_resid_post

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

32,768

Data Type

float32

Hook Name

blocks.12.hook_resid_post

Architecture

jumprelu

Context Size

1,024

Dataset

chanind/pile-uncopyrighted-gemma-1024-abbrv-1B

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

KommentareTeilen

-0.80

unmodifiable

-0.77

SPY

-0.76

ividual

-0.75

PLN

-0.75

氓

-0.75

olesome

-0.75

 dran

-0.75

StructEnd

-0.74

dshaw

-0.73

POSITIVE LOGITS

 certificate

1.35

 certificates

1.34

cer

1.34

 cert

1.28

 CERT

1.25

 certifying

1.25

Cer

1.23

Cer

1.23

 Certificate

1.23

 Cert

1.23

Activations Density 0.066%

terms related to legal certifications and court rulings

No Comments

No Known Activations