INDEX

Explanations

punctuation marks, specifically periods

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

programming language keywords

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

 betweenstory

-0.47

 asimismo

-0.44

ActionCreators

-0.43

webElementXpaths

-0.41

SOUNDBITE

-0.40

 ModelExpression

-0.39

AnimationsModule

-0.39

 HasFactory

-0.38

 inoltre

-0.38

 validamos

-0.36

POSITIVE LOGITS

󠁣

0.62

queryInterface

0.54

OGND

0.53

########.

0.50

glBind

0.50

 levy

0.48

talt

0.47

taler

0.47

millimeters

0.47

kháu

0.47

Activations Density 0.001%

punctuation marks, specifically periods

programming language keywords

No Comments

No Known Activations