INDEX

Explanations

expressions of common knowledge and familiarity with subjects or cultural references

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 21-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_21/width_16k/average_l0_129

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.21.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

 запи

-0.30

PREFERRED

-0.30

 Neel

-0.30

vit

-0.30

Tink

-0.29

SFD

-0.29

Dusty

-0.28

sonian

-0.28

 getContent

-0.28

 getPath

-0.27

POSITIVE LOGITS

 familiar

0.86

 familiarity

0.71

familiar

0.68

Familiar

0.68

 Familiar

0.66

 bekan

0.65

 disambiguazione

0.65

OGND

0.65

 famous

0.64

 known

0.64

Activations Density 0.265%

expressions of common knowledge and familiarity with subjects or cultural references

No Comments

No Known Activations