INDEX

Explanations

references to specific names or titles

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

disambiguation pages

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

AndEndTag

-0.47

heda

-0.38

 Türkei

-0.37

Pä

-0.37

Preparación

-0.36

chetto

-0.36

grafías

-0.36

当即

-0.36

vs

-0.36

cellation

-0.36

POSITIVE LOGITS

 disambiguazione

0.65

 ujednoznacz

0.65

contentLoaded

0.63

verwijspagina

0.53

 transfieras

0.52

rungsseite

0.46

曖昧さ回避

0.45



0.44

ecma

0.43

 Normdatei

0.42

Activations Density 0.003%

references to specific names or titles

disambiguation pages

No Comments

No Known Activations