INDEX

Explanations

actions typically associated with browsing or engaging in online activities

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

user wanting something

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

parsedMessage

-0.50

épend

-0.47

 påver

-0.46

KURZBESCHREIBUNG

-0.44

évaluateur

-0.44

 bluzka

-0.43

optionalTypeArgs

-0.43

tanleria

-0.42

 nakalista

-0.42

endregion

-0.41

POSITIVE LOGITS

olin

0.45

 NSCoder

0.43

 Stella

0.41

DockStyle

0.41

 bowl

0.41

folger

0.40

 بيها

0.40

onda

0.40

lini

0.39

OGND

0.39

Activations Density 0.034%

actions typically associated with browsing or engaging in online activities

user wanting something

No Comments

No Known Activations