INDEX

Explanations

phrases containing the word "able."

oai_token-act-pair · gpt-3.5-turbo

phrases indicating capability or ability

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

ccording

-0.59

bour

-0.53

 Bots

-0.53

 Nose

-0.51

 Deer

-0.48

afort

-0.48

THING

-0.48

 Decay

-0.47

 Parents

-0.47

 Gone

-0.47

POSITIVE LOGITS

to

1.16

bodied

1.04

 thereto

0.82

to

0.81

ta

0.77

ioned

0.77

Reviewer

0.74

bod

0.72

iary

0.71

TO

0.71

Activations Density 0.060%

phrases containing the word "able."

phrases indicating capability or ability

No Comments

No Known Activations