INDEX

Explanations

specific brand names, categories, and classifications within various contexts

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 <=",

-0.67

WebElementEntity

-0.63

webElementXpaths

-0.56

IntoConstraints

-0.49

featureID

-0.48

principalColumn

-0.47

AutoField

-0.46



-0.46

toMatchSnapshot

-0.45

 typelib

-0.45

POSITIVE LOGITS

 pavillon

0.43

fromnode

0.43

 wiki

0.42

 <<<<<<<<<<<<<<

0.41

Chham

0.40

wiki

0.40

map

0.40

playerName

0.40

profession

0.40

toolStripButton

0.40

Activations Density 0.014%

specific brand names, categories, and classifications within various contexts

No Comments

No Known Activations