INDEX

Explanations

contractions of "we are" or "I am"

oai_token-act-pair · gpt-3.5-turbo

the contraction and use of the words "we're" and "I'm" to indicate states of being or feeling

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

 mater

-0.78

={

-0.64

 Reduce

-0.63

-0.62

 separates

-0.61

ESE

-0.59

 membr

-0.58

andise

-0.58

TAIN

-0.58

POSITIVE LOGITS

 gonna

1.35

 gotta

1.06

 hoping

0.91

 supposed

0.87

 going

0.85

 sorry

0.85

got

0.83

 glad

0.82

 lucky

0.81

 afraid

0.81

Activations Density 0.067%

contractions of "we are" or "I am"

the contraction and use of the words "we're" and "I'm" to indicate states of being or feeling

No Comments

No Known Activations