INDEX

Explanations

mentions of governmental agencies or organizations

oai_token-act-pair · gpt-3.5-turbo

references to a specific agency or organization

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.2.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Dear

-0.70

Tea

-0.70

lihood

-0.69

aston

-0.68

 Shields

-0.65

âķĲâķĲ

-0.64

Klu

-0.64

nder

-0.63

 Bened

-0.63

isers

-0.63

POSITIVE LOGITS

 agency

1.11

 agencies

1.02

encies

0.88

 administrator

0.81

acco

0.79

 contractor

0.78

 Agency

0.77

 director

0.76

ENCY

0.76

 regulator

0.74

Activations Density 0.014%

mentions of governmental agencies or organizations

references to a specific agency or organization

No Comments

No Known Activations