INDEX

Explanations

names of individuals

oai_token-act-pair · gpt-3.5-turbo

sequences of names or titles

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 redes

-0.85

 favour

-0.84

rul

-0.83

 Unch

-0.82

 crumble

-0.81

 snowball

-0.80

 yield

-0.79

 proble

-0.76

 spir

-0.75

 resolved

-0.75

POSITIVE LOGITS

Mike

1.94

Jason

1.90

Larry

1.90

Michael

1.90

Chris

1.89

Richard

1.88

Brian

1.88

Gary

1.88

Kevin

1.88

David

1.86

Activations Density 0.141%

names of individuals

sequences of names or titles

No Comments

No Known Activations