INDEX

Explanations

terms related to application features and user interactions

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

specific technical or foreign words

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

pyplot

-0.43

InteropServices

-0.34

PropertyChanging

-0.33

 Administrativna

-0.33

 ivelany

-0.32

mi

-0.31

 belang

-0.30

stdc

-0.30

Naissance

-0.30

曖昧さ回避

-0.30

POSITIVE LOGITS

 BoxFit

0.64

 Autorizaciones

0.60

InjectAttribute

0.56

 Autorisations

0.56

 ſeinen

0.56

évaluateur

0.54

Efq

0.52

 ſeine

0.52

RTEE

0.52

AsUp

0.51

Activations Density 0.124%

terms related to application features and user interactions

specific technical or foreign words

No Comments

No Known Activations