INDEX

Explanations

occurrences and definitions of class-related keywords in programming

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

class name or method call

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

 المعيارى

-0.48

SharedCtor

-0.47

 <=",

-0.45

AndEndTag

-0.45

########.

-0.44

plotlib

-0.43

 ब्रेकडाउन

-0.42

ftagPool

-0.41

 seier

-0.40

 afges

-0.40

POSITIVE LOGITS

NewGuid

0.49

 consultato

0.42

printStackTrace

0.42

 ویکی‌پدیا

0.41

Spoljašnje

0.40

RTSN

0.39

évaluateur

0.38

osoba

0.36

 sapiens

0.36

RuntimeObject

0.36

Activations Density 0.004%

occurrences and definitions of class-related keywords in programming

class name or method call

No Comments

No Known Activations