INDEX

Explanations

hints or suggestions within a text

oai_token-act-pair · gpt-3.5-turbo

indications or suggestions of underlying themes or meanings in the text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

ocker

-0.88

nea

-0.80

ccording

-0.78

ctic

-0.75

artney

-0.74

ridges

-0.72

ameron

-0.69

lim

-0.68

ruciating

-0.66

animate

-0.66

POSITIVE LOGITS

 hint

1.64

 hints

1.38

 clue

0.93

 wink

0.80

 clues

0.77

 hinted

0.76

 Leilan

0.76

llor

0.75

 warning

0.74

TextColor

0.71

Activations Density 0.007%

hints or suggestions within a text

indications or suggestions of underlying themes or meanings in the text

No Comments

No Known Activations