INDEX

Explanations

phrases related to actions and events

oai_token-act-pair · gpt-3.5-turbo

conjunctions and phrases indicating continuation in sentences

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Kills

-0.76

rush

-0.72

 Digest

-0.68

track

-0.63

use

-0.58

mix

-0.58

draft

-0.57

 Manifest

-0.56

 Rapid

-0.55

 Supply

-0.54

POSITIVE LOGITS

 gotten

1.76

been

1.70

 been

1.62

 gone

1.55

 eaten

1.52

 taken

1.48

 begun

1.47

 ridden

1.41

 done

1.40

 risen

1.38

Activations Density 0.196%

phrases related to actions and events

conjunctions and phrases indicating continuation in sentences

No Comments

No Known Activations