INDEX

Explanations

the word "By" and its variations, indicating a focus on authorship or agency

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 itſelf

-0.71

 ſtate

-0.68

 ſal

-0.65

 leaſt

-0.65

 myſelf

-0.64

 Waſſer

-0.64

ésultats

-0.63

 feroit

-0.61

iſter

-0.60

leſs

-0.60

POSITIVE LOGITS

by

1.19

by

1.13

BY

1.09

By

1.09

By

1.08

BY

1.05

 getBy

0.90

 Byers

0.85

bys

0.79

getBy

0.79

Activations Density 0.016%

the word "By" and its variations, indicating a focus on authorship or agency

No Comments

No Known Activations