INDEX

Explanations

hints or suggestions within a text

oai_token-act-pair · gpt-3.5-turbo

suggestions or indications of future developments or outcomes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

vict

-0.72

ÄŁ

-0.71

ruciating

-0.69

exper

-0.69

rament

-0.69

lux

-0.66

Exper

-0.62

zanne

-0.61

rior

-0.61

 seiz

-0.60

POSITIVE LOGITS

 hint

1.41

 hints

1.28

 clue

1.04

 clues

1.00

 hinted

0.97

 glimps

0.88

 indicating

0.75

ule

0.72

 towards

0.72

 wink

0.71

Activations Density 0.068%

hints or suggestions within a text

suggestions or indications of future developments or outcomes

No Comments

No Known Activations