INDEX

Explanations

elements related to the act of communication or conversational context

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

statistical and unusual 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

 AspNetCore

-0.50

archiviato

-0.47

 tartalomajánló

-0.43

gentes

-0.42

 getItemId

-0.42

 مواليد

-0.41

 tolua

-0.40

INCREF

-0.40

日閲覧

-0.39

 hava

-0.39

POSITIVE LOGITS

[toxicity=0]

0.77

GEBURTSDATUM

0.73

 leſs

0.59

Diweddarwch

0.58

 itſelf

0.57

 houſe

0.56



0.56

ſelves

0.56

Tikang

0.54

 faſt

0.54

Activations Density 0.068%

elements related to the act of communication or conversational context

statistical and unusual words

No Comments

No Known Activations