INDEX

Explanations

punctuation marks such as commas and semicolons

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

Keywords, abbreviations, identifiers

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

LMS

-0.43

ariamente

-0.42

WireFormatLite

-0.39

czegó

-0.38

sch

-0.38

 ebenfalls

-0.37

 jist

-0.37

UserScript

-0.37

EHR

-0.36

GEM

-0.36

POSITIVE LOGITS

Personendaten

0.75

GEBURTS

0.65

 transfieras

0.56

 تضيفلها

0.55

 Савезне

0.54

Gön

0.50

ragalactic

0.50

KURZBESCHREIBUNG

0.47

Vidite

0.47

utilisons

0.46

Activations Density 0.004%

punctuation marks such as commas and semicolons

Keywords, abbreviations, identifiers

No Comments

No Known Activations