INDEX

Explanations

reported speech or dialogue in the text

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

reporting verbs and speakers

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

ſelf

-0.70

LLocation

-0.68

 ferons

-0.64

ſelves

-0.62

 poffe

-0.56

 queſta

-0.56

 financières

-0.53

 ſever

-0.53

 nationales

-0.53

 ſhall

-0.53

POSITIVE LOGITS

帖最后由

0.54

GOTREF

0.54

Mr

0.45

 חיצוניים

0.45

the

0.44

://$

0.44

0.43

by

0.42

這位

0.42

 oleh

0.40

Activations Density 0.006%

reported speech or dialogue in the text

reporting verbs and speakers

No Comments

No Known Activations