INDEX

Explanations

references to authors or writers

oai_token-act-pair · gpt-3.5-turbo

mentions of authors in the text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

EMS

-0.71

 Nicaragua

-0.66

tone

-0.66

ll

-0.65

ering

-0.62

 Rite

-0.62

 Territories

-0.62

 Buenos

-0.62

ndum

-0.62

yip

-0.61

POSITIVE LOGITS

itatively

1.74

itar

1.38

itative

1.29

itarian

1.26

hip

1.00

isations

0.97

izations

0.96

izes

0.88

sonian

0.88

isation

0.87

Activations Density 0.036%

references to authors or writers

mentions of authors in the text

No Comments

No Known Activations