INDEX

Explanations

references to authority or specific entities like organizations or people

oai_token-act-pair · gpt-3.5-turbo

the presence of the word "the"

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

ategory

-0.87

thood

-0.84

ictionary

-0.80

Alert

-0.79

CHAPTER

-0.79

Serv

-0.78

bg

-0.76

ustom

-0.76

icia

-0.75

aternity

-0.73

POSITIVE LOGITS

 slightest

1.11

 feds

1.06

 majority

1.04

 Russians

0.95

 Soviets

0.94

 entire

0.92

 vast

0.89

 answer

0.88

guy

0.88

 Clintons

0.86

Activations Density 0.347%

references to authority or specific entities like organizations or people

the presence of the word "the"

No Comments

No Known Activations