INDEX

Explanations

code-related terms and constructs

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

code assignments and structure

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

 otomatig

-0.74

 Wiktionnaire

-0.66

 فريبيس

-0.64

 مشين

-0.60

addGap

-0.59

NameInMap

-0.58

GEBURTS

-0.58

 تضيفلها

-0.57

 disambiguazione

-0.54

verwijspagina

-0.54

POSITIVE LOGITS

fsch

0.34

owią

0.33

aR

0.32

hafen

0.31

 bornes

0.31

CD

0.31

 sobremesa

0.30

임

0.30

 verstanden

0.30

 saksi

0.30

Activations Density 0.013%

code-related terms and constructs

code assignments and structure

No Comments

No Known Activations