INDEX

Explanations

phrases indicating a declaration or presentation of information

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

declared that

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

ьажоргаш

-0.46

 Dirig

-0.41

awtextra

-0.40

 Chiara

-0.38

 nød

-0.36

 Bulgarian

-0.36

 Unwin

-0.36

stim

-0.35

 scha

-0.35

 stim

-0.35

POSITIVE LOGITS

WriteBarrier

0.56

 pinulongan

0.54

 AssemblyTitle

0.50

__":

0.46

ValueStyle

0.46

windowFixed

0.44

Instrucciones

0.44

ApplicationTests

0.44

__":

0.43

AndEndTag

0.43

Activations Density 0.002%

phrases indicating a declaration or presentation of information

declared that

No Comments

No Known Activations