INDEX

Explanations

hints or suggestions

oai_token-act-pair · gpt-3.5-turbo

words that indicate subtle suggestions or clues

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

nea

-0.78

ccording

-0.76

ocker

-0.71

portion

-0.69

vict

-0.69

frey

-0.68

bred

-0.66

CENT

-0.64

bel

-0.63

ctic

-0.63

POSITIVE LOGITS

 hint

1.48

 hints

1.46

 clues

0.88

 hinted

0.87

 clue

0.85

 wink

0.80

itives

0.80

 glimps

0.77

 warning

0.76

llor

0.73

Activations Density 0.018%

hints or suggestions

words that indicate subtle suggestions or clues

No Comments

No Known Activations