INDEX

Explanations

references to specific individuals and their actions

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

names and technical terms

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

 productive

-0.35

Rifer

-0.35

 sper

-0.34

EP

-0.34

gı

-0.33

ongi

-0.32

 aDecoder

-0.32

files

-0.32

мен

-0.31

 Mack

-0.31

POSITIVE LOGITS

InjectAttribute

0.68

richTextPanel

0.66

findpost

0.63

oa̍t

0.62

twimg

0.56

parsedMessage

0.56

 estekak

0.56

 transférez

0.54

ंदीखरीदारी

0.54

msgTypes

0.54

Activations Density 0.297%

references to specific individuals and their actions

names and technical terms

No Comments

No Known Activations