INDEX

Explanations

introductions and transitions in written text

oai_token-act-pair · gpt-3.5-turbo

indications of introductory or preparatory statements

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 gull

-0.66

nell

-0.63

 quarry

-0.63

arov

-0.63

 Towns

-0.62

 lining

-0.61

ussia

-0.61

alties

-0.61

mons

-0.60

lot

-0.60

POSITIVE LOGITS

 summarize

1.00

 recap

0.93

 briefly

0.93

 summar

0.93

 disclaimer

0.90

 reiterate

0.88

bre

0.84

 spoiler

0.84

Alright

0.83

claimer

0.82

Activations Density 0.999%

introductions and transitions in written text

indications of introductory or preparatory statements

No Comments

No Known Activations