INDEX

Explanations

programming-related terminologies and conditional statement endings

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 0-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_0/width_65k/average_l0_73

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.0.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

 myſelf

-0.81

 AssemblyCompany

-0.61

 fhort

-0.60

 auffi

-0.60

 himſelf

-0.59

 ainfi

-0.58

 dieß

-0.57

 perus

-0.57

 purpoſe

-0.57

 precis

-0.56

POSITIVE LOGITS

fel

0.66

 scanf

0.62

endif

0.57

 bParam

0.56

feldt

0.56

Fel

0.54

ishman

0.54

ometrica

0.53

wanne

0.52

лер

0.52

Activations Density 0.039%

programming-related terminologies and conditional statement endings

No Comments

No Known Activations