INDEX

Explanations

phrases and expressions of meaning or intention

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

does not mean

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

AndEndTag

-0.56

 gewor

-0.46

RenderAtEndOf

-0.45

 bookId

-0.45

énario

-0.45

 projectId

-0.43

 ویکی‌آمباردا

-0.43

 pungkasnya

-0.42

 écou

-0.42

 BigNumberish

-0.42

POSITIVE LOGITS

 Middlesex

0.42

stateProvider

0.42

 future

0.40

 OMIT

0.40

 Auburn

0.39

replay

0.37

ChildIndex

0.37

 Isaiah

0.37

 Chesapeake

0.37

 agents

0.36

Activations Density 0.013%

phrases and expressions of meaning or intention

does not mean

No Comments

No Known Activations