INDEX

Explanations

phrases indicating liability or accountability

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 12-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_12/width_16k/average_l0_82

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.12.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 lovely

-0.59

 nice

-0.56

î

-0.55

we

-0.55

 nicer

-0.55

-0.54

 pretty

-0.54

mathrm

-0.53

 maman

-0.53

POSITIVE LOGITS

findpost

0.98

Personensuche

0.87

 prêtres

0.85

出版年

0.83

 stället

0.82

ronpa

0.80

seamnă

0.80

 préférence

0.79

 dezelve

0.79

 varandra

0.79

Activations Density 1.120%

phrases indicating liability or accountability

No Comments

No Known Activations