INDEX

Explanations

dates and significant years in historical contexts

oai_token-act-pair · gpt-3.5-turbo

sentence endings or conclusions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

 selfie

-0.79

pit

-0.77

 pudding

-0.76

inki

-0.76

 emoji

-0.75

 neigh

-0.75

 vape

-0.74

 minion

-0.73

 purse

-0.72

arse

-0.72

POSITIVE LOGITS

 Eventually

1.48

 Later

1.42

 Afterwards

1.39

 Shortly

1.29

 Ultimately

1.29

 Then

1.25

 Unfortunately

1.22

 Soon

1.21

 During

1.20

 Ironically

1.18

Activations Density 0.593%

dates and significant years in historical contexts

sentence endings or conclusions

No Comments

No Known Activations