INDEX

Explanations

code snippets and programming-related terminology

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 39-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_39/width_16k/average_l0_131

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.39.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

into

-0.52

typeorm

-0.50

elemField

-0.50

olism

-0.47

Tembelea

-0.47

 Into

-0.47

webElementGuid

-0.44

Into

-0.44

données

-0.44

tangentMode

-0.43

POSITIVE LOGITS

 Geſ

0.53

 argint

0.49

 entanto

0.48

 Bucure

0.47

 cuad

0.47

 нѣ

0.46

 зовут

0.46

 Wikiseite

0.45

 rechange

0.45

この日は

0.45

Activations Density 1.821%

code snippets and programming-related terminology

No Known Activations