INDEX

Explanations

references to the word "thing."

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

the word thing

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 31-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_31/width_131k/average_l0_109

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.31.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

 }}"></

-0.54

Autoritní

-0.54

 EconPapers

-0.53

GIVEREF

-0.53

 Савезне

-0.53

GEBURTSDATUM

-0.52

OGND

-0.52

%"),

-0.52

__":

-0.51

GOTREF

-0.51

POSITIVE LOGITS

 itself

0.53

ik

0.51

ie

0.51

ist

0.49

LookAnd

0.47

ning

0.47

ier

0.46

ip

0.44

spot

0.44

 Drapeau

0.42

Activations Density 0.012%

references to the word "thing."

the word thing

No Comments

No Known Activations