INDEX

Explanations

titles and positions related to authority or leadership roles

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

titles of people or positions

np_acts-logits-general · gemini-2.5-flash-lite

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

featureID

-0.60

 postIndex

-0.54

WriteAttribute

-0.50

msgTypes

-0.47

MigrationBuilder

-0.47

طلحات

-0.43

äischen

-0.43

printStackTrace

-0.42

moveToFirst

-0.42

Lef

-0.41

POSITIVE LOGITS

httphttps

0.52

 createSlice

0.47

}{@

0.46

ReactDOM

0.42

WebpackPlugin

0.41

(!__

0.40

 łą

0.39

 AppComponent

0.38

RunWith

0.38

arote

0.36

Activations Density 0.027%

titles and positions related to authority or leadership roles

titles of people or positions

No Comments

No Known Activations