INDEX

Explanations

code and programming-related terms

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 13-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_13/width_65k/average_l0_75

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.13.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

encodeWith

-0.55

 signin

-0.52

loadAll

-0.50

 abzu

-0.50

ecuzione

-0.50

bonucleic

-0.48

 addTo

-0.47

 alqu

-0.46

 dissip

-0.46

 speculative

-0.46

POSITIVE LOGITS

 filter

0.71

 Filter

0.66

 remove

0.64

 Remove

0.61

 FILTER

0.59

 convert

0.59

filter

0.59

 conversion

0.59

Filter

0.58

 jadx

0.57

Activations Density 0.757%

code and programming-related terms

No Comments

No Known Activations