INDEX

Explanations

positive adjectives like "awesome" with moderately high activation values

oai_token-act-pair · gpt-3.5-turbo

instances of the word "awesome" and its variations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

mediate

-0.70

Clar

-0.68

idation

-0.66

reditation

-0.65

Downloadha

-0.64

apers

-0.63

avis

-0.63

 Inquiry

-0.63

licts

-0.63

ASE

-0.63

POSITIVE LOGITS

ly

0.98

ery

0.97

ness

0.93

 stuff

0.82

NESS

0.80

stuff

0.77

 artwork

0.77

nels

0.76

GIF

0.76

 sounding

0.75

Activations Density 0.075%

positive adjectives like "awesome" with moderately high activation values

instances of the word "awesome" and its variations

No Comments

No Known Activations