INDEX

Explanations

references to filenames and timestamps in code or data processing contexts

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

programming data 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

getUsername

-0.61

 المعيارى

-0.55

stdc

-0.54

AnimationsModule

-0.52

OGND

-0.52

 hierv

-0.51

 setUsername

-0.49

 saç

-0.48

Myth

-0.48

ísmo

-0.48

POSITIVE LOGITS

 userName

0.63

 DataBase

0.63

 callBack

0.62

er

0.61

uxxxx

0.61

callBack

0.60

toDo

0.58

UserName

0.57

 propOrder

0.57

DataBase

0.56

Activations Density 0.009%

references to filenames and timestamps in code or data processing contexts

programming data types

No Comments

No Known Activations