INDEX

Explanations

programming-related keywords and code comments

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

code blocks and keywords

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

worms

-0.43

worm

-0.43

 Neop

-0.41

 Interna

-0.41

emias

-0.41

 Heid

-0.41

cocks

-0.41

eby

-0.38

 campan

-0.38

 sall

-0.38

POSITIVE LOGITS

DockStyle

0.58

 estekak

0.54

ResumeLayout

0.54

تقاوى

0.53

 NSCoder

0.52

 chi̍t

0.50

 wikipagina

0.50

IVEREF

0.50

 ویکی‌پدیا

0.50

RTEE

0.48

Activations Density 0.016%

programming-related keywords and code comments

code blocks and keywords

No Comments

No Known Activations