INDEX

Explanations

phrases indicating missed opportunities or things that could have been done differently

oai_token-act-pair · gpt-3.5-turbo

the phrase "could have" indicating unrealized potential or missed opportunities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

 paycheck

-0.65

 craving

-0.61

Cly

-0.60

 hostage

-0.59

stown

-0.57

Mem

-0.56

 doll

-0.55

 reminder

-0.53

 grappling

-0.53

shy

-0.53

POSITIVE LOGITS

ħĭ

0.97

 been

0.89

 Been

0.83

 gotten

0.82

ŃĶ

0.80

idon

0.79

¶ħ

0.79

ivably

0.78

been

0.75

Ģ

0.73

Activations Density 0.051%

phrases indicating missed opportunities or things that could have been done differently

the phrase "could have" indicating unrealized potential or missed opportunities

No Comments

No Known Activations