INDEX

Explanations

references to high-level job titles and roles within organizations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 10-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_10/width_16k/average_l0_77

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.10.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

 intérêt

-0.66

INCLUDED

-0.64

toHaveBeenCalled

-0.64

ấc

-0.62

cipl

-0.56

 Trabalho

-0.56

 Tasse

-0.56

Bibliograf

-0.53

 Specialization

-0.53

ensatz

-0.52

POSITIVE LOGITS

CEO

0.92

CEO

0.80

ceo

0.78

 CEOs

0.76

VersionUID

0.68

contentLoaded

0.66

BeginContext

0.60

MockBean

0.59

AntiForgeryToken

0.59

Activations Density 0.076%

references to high-level job titles and roles within organizations

No Comments

No Known Activations