INDEX

Explanations

conversational cues and questions indicating uncertainty or curiosity

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

foreigners and proper nouns

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

 ſind

-0.72

ſelf

-0.71

AsUp

-0.71

ſelves

-0.65

 '\\;'

-0.61

mobileqq

-0.60

BibitemShut

-0.60

RTLR

-0.57

niſſe

-0.57

ſicht

-0.56

POSITIVE LOGITS

:✨

0.40

 fuck

0.32

 templado

0.32

 Hauser

0.32

 foreigners

0.30

 major

0.30

 fédé

0.30

 Silver

0.30

 silver

0.29

ここまで

0.29

Activations Density 0.043%

conversational cues and questions indicating uncertainty or curiosity

foreigners and proper nouns

No Comments

No Known Activations