INDEX

Explanations

non-informative or filler tokens in the text

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

version numbers and code

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

ImageContext

-0.46

DataSetChanged

-0.45

extré

-0.43

DIPSETTING

-0.43

setPointSize

-0.40

 ceinture

-0.40

esomeness

-0.40

 considération

-0.40

ExtendWith

-0.40

ALIAS

-0.39

POSITIVE LOGITS

ScopeManager

0.54

httphttps

0.51

uxxxx

0.48

المناصب

0.46

Tembelea

0.45

DeleteBehavior

0.44

şiv

0.44

 CallOverrides

0.43

 comuniques

0.42

 protoimpl

0.41

Activations Density 0.000%

non-informative or filler tokens in the text

version numbers and code

No Comments

No Known Activations