INDEX

Explanations

programming-related class and method structures

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

code identifiers and types

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

enterOuterAlt

-0.38

LogFactory

-0.34

 sumpay

-0.34

InSection

-0.33

ArgsConstructor

-0.31

 Italijani

-0.31

AntiForgeryToken

-0.30

GraphicsUnit

-0.30

KommentareTeilen

-0.30

DoubleQuotes

-0.30

POSITIVE LOGITS

bootstrapcdn

0.59

Tikang

0.56

ſelves

0.56

iastical

0.54

 iſt

0.54

thentication

0.53

mnop

0.53

zzang

0.53

formable

0.52

philosop

0.52

Activations Density 0.039%

programming-related class and method structures

code identifiers and types

No Comments

No Known Activations