INDEX

Explanations

images or photos mentioned in a text

oai_token-act-pair · gpt-3.5-turbo

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ItemTracker

-1.02

wcs

-0.79

few

-0.78

acious

-0.77

rss

-0.76

endiary

-0.76

acies

-0.75

asons

-0.73

WAYS

-0.73

acist

-0.72

POSITIVE LOGITS

 flowers

0.93

 smiling

0.90

tnc

0.86

him

0.86

 naked

0.85

 oneself

0.84

 themselves

0.80

 kittens

0.80

 himself

0.78

 Himself

0.77

Activations Density 0.166%

images or photos mentioned in a text

No Comments

No Known Activations