INDEX

Explanations

mentions or honors

oai_token-act-pair · gpt-3.5-turbo

instances of the term "honorable mention" in various contexts

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

 licences

-0.72

 devote

-0.72

 leaf

-0.65

 clubs

-0.64

 flags

-0.62

 advertising

-0.60

 cricket

-0.59

 flats

-0.58

 licence

-0.58

 cross

-0.57

POSITIVE LOGITS

orable

4.73

oring

1.47

ored

1.46

ORED

1.31

orage

1.26

orious

1.22

oration

1.19

orer

1.18

urable

1.14

orative

1.13

Activations Density 0.008%

mentions or honors

instances of the term "honorable mention" in various contexts

No Comments

No Known Activations