INDEX

Explanations

phrases indicating external sources of information or hearsay

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

seeking to understand or learn

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

IntoConstraints

-0.64

 informée

-0.51

Демографія

-0.51

Personensuche

-0.48

SharedCtor

-0.48

 NSCoder

-0.48

 ExecuteAsync

-0.48

:✨

-0.47

typeorm

-0.45

 newOwner

-0.44

POSITIVE LOGITS

uttgart

0.42

有的

0.39

 xmlns

0.39

 dikenal

0.37

makeText

0.36

loween

0.35

听说

0.35

fopen

0.35

 important

0.35

اقرأ

0.35

Activations Density 0.041%

phrases indicating external sources of information or hearsay

seeking to understand or learn

No Comments

No Known Activations