INDEX

Explanations

specific entities, places, or names of people and their relevance in the context

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

own/proper 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



-0.84

 transfieras

-0.70

 للمعارف

-0.68

TypedDataSet

-0.68

دانشنامهٔ

-0.65

RenderAtEndOf

-0.65

UrlResolution

-0.65

 Normdatei

-0.64

 esternos

-0.64

 Италијани

-0.63

POSITIVE LOGITS

 propre

0.34

̩

0.32

 menge

0.30

ó

0.30

sí

0.30

 próprios

0.29

 quebra

0.28

́

0.27

 estre

0.27

 pard

0.27

Activations Density 0.519%

specific entities, places, or names of people and their relevance in the context

own/proper words

No Comments

No Known Activations