INDEX

Explanations

words related to specific locations, such as "Oswalt" and "Islington."

oai_token-act-pair · gpt-3.5-turbo

references to specific individuals and locations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 Frozen

-0.77

 Illusion

-0.76

 Masquerade

-0.75

 Apocalypse

-0.73

 Hawaiian

-0.72

 vape

-0.71

ĸļ

-0.70

 Cinderella

-0.70

illac

-0.69

 Crunch

-0.68

POSITIVE LOGITS

Osw

2.90

lington

1.53

orthy

1.51

Yar

1.37

 Berman

1.29

abilia

1.24

 Rutherford

1.10

 Thames

1.09

 leftist

1.02

 Torres

0.98

Activations Density 0.062%

words related to specific locations, such as "Oswalt" and "Islington."

references to specific individuals and locations

No Comments

No Known Activations