INDEX

Explanations

nouns and their associated values or identifiers

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

money, activity, therapy, vision

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

PostBack

-0.37

 esternos

-0.34

enoord

-0.34

 NSCoder

-0.34

AndEndTag

-0.32

webElementGuid

-0.32

sbury

-0.32

 endwhile

-0.31

 belo

-0.30

utr

-0.30

POSITIVE LOGITS

ⓧ

0.64

 pinulongan

0.62

UserScript

0.58

styleType

0.55

SharedCtor

0.54

providedIn

0.52

 تانيه

0.51

rrggbb

0.50

+#+#

0.48

SourceChecksum

0.48

Activations Density 0.072%

nouns and their associated values or identifiers

money, activity, therapy, vision

No Comments

No Known Activations