INDEX

Explanations

references to locations, specifically the state of Ohio

oai_token-act-pair · gpt-3.5-turbo

references to the state of Ohio

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

 learnt

-0.78

 dragons

-0.74

tel

-0.71

 mages

-0.71

 Mages

-0.70

 realised

-0.68

 teleport

-0.68

 realise

-0.67

 cafÃ©

-0.66

MAL

-0.66

POSITIVE LOGITS

 Ohio

3.66

Ohio

3.39

 Akron

2.08

 Michigan

2.06

Michigan

2.00

 Indiana

1.98

 Illinois

1.98

 Pennsylvania

1.97

 Missouri

1.91

 Cleveland

1.88

Activations Density 0.031%

references to locations, specifically the state of Ohio

references to the state of Ohio

No Comments

No Known Activations