INDEX

Explanations

references to examples or illustrations within a text

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

law

-0.38

Liefs

-0.37

 familiar

-0.33

(""));

-0.33

 central

-0.33

pecabezas

-0.32

krieg

-0.32

Übersetzung

-0.32

 kawaida

-0.31

POSITIVE LOGITS

 autorytatywna

0.66

 estekak

0.59

 surla

0.57

Suara

0.55

évaluateur

0.54

ThroughAttribute

0.54

 لينك

0.53

帖最后由

0.51

delwed

0.50

NewUrlParser

0.50

Activations Density 0.007%

references to examples or illustrations within a text

No Comments

No Known Activations