INDEX

Explanations

specific nouns and terms related to biology and anatomical structures

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

specific object parts

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

 fört

-0.32

̈́

-0.31

 terbaru

-0.31

€“

-0.30

 taglia

-0.29

hence

-0.29

Referanser

-0.28

 kohta

-0.28

 hoog

-0.28

 fær

-0.28

POSITIVE LOGITS

titleMargin

0.65

bcryptjs

0.64

lapsingToolbar

0.59

 UIAlert

0.56

LLocation

0.56

AnchorStyles

0.56

 miniaturka

0.56

ſelf

0.56

windowFixed

0.54

 BioLib

0.53

Activations Density 0.164%

specific nouns and terms related to biology and anatomical structures

specific object parts

No Comments

No Known Activations