INDEX

Explanations

numbers or numerical expressions

oai_token-act-pair · gpt-3.5-turbo

the end of a document or text segment

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

schild

-0.72

bies

-0.68

 vanity

-0.65

 bucket

-0.65

 treasury

-0.64

 charm

-0.64

 footing

-0.63

 fronts

-0.63

 bait

-0.62

 fert

-0.62

POSITIVE LOGITS

 ILCS

1.09

 Immun

0.76

 largeDownload

0.73

NetMessage

0.73

=>

0.73

 Frames

0.72

 DISTRICT

0.70

 Retrieved

0.69

 Unloaded

0.69

ACE

0.69

Activations Density 0.029%

numbers or numerical expressions

the end of a document or text segment

No Comments

No Known Activations