INDEX

Explanations

keywords related to object-oriented programming constructs such as classes, methods, and overrides

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

override keyword

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

Personendaten

-0.72

makeText

-0.71

 BoxFit

-0.67

ScopeManager

-0.59

 HFILL

-0.56

SequentialGroup

-0.54

UnknownFieldSet

-0.54

 MainAxisSize

-0.54

Fordítás

-0.53

oredCriteria

-0.53

POSITIVE LOGITS

NameInMap

0.36

 home

0.35

cidae

0.33

mui

0.32

vello

0.32

jabat

0.32

代

0.32

GOB

0.31

 Douglass

0.31

 אותו

0.31

Activations Density 0.009%

keywords related to object-oriented programming constructs such as classes, methods, and overrides

override keyword

No Comments

No Known Activations