INDEX

Explanations

references to various types of "other" factors or issues

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

other and list items

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

ंदीखरीदारी

-0.41

tonode

-0.40

 verrez

-0.39

chargez

-0.39

الدراسه

-0.38

PerformLayout

-0.37

seitig

-0.37

 TAMBÉM

-0.37

iseries

-0.37

brä

-0.36

POSITIVE LOGITS

lichter

0.47

 sigui

0.42

ніципалі

0.41

etc

0.41

Etc

0.41

EReference

0.41

 všetko

0.39

何より

0.39

Kariera

0.39

ETC

0.39

Activations Density 0.026%

references to various types of "other" factors or issues

other and list items

No Comments

No Known Activations