INDEX

Explanations

URLs or identifiers typically associated with citations or web resources

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

URLs with IDs

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

Diwedd

-0.50

InputBorder

-0.47

 Winaray

-0.46

⟬

-0.46

 المعيارى

-0.46

 €€

-0.43

eur

-0.42

ANDUM

-0.42

 ſch

-0.40

mnop

-0.40

POSITIVE LOGITS



0.69

AndEndTag

0.68

findpost

0.54

Portail

0.51

utilisons

0.45

 maîtres

0.43

zeiger

0.43

Посилання

0.42

stateProvider

0.42

 transfieras

0.41

Activations Density 0.006%

URLs or identifiers typically associated with citations or web resources

URLs with IDs

No Comments

No Known Activations