INDEX

Explanations

phrases related to problem-solving or fixing issues

oai_token-act-pair · gpt-3.5-turbo

instances of the word "fix" and its variations

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

NetMessage

-0.75

ographies

-0.71

ographic

-0.69

Champ

-0.65

ammy

-0.63

 Schwar

-0.63

DragonMagazine

-0.62

Rush

-0.61

SPONSORED

-0.61

Sak

-0.61

POSITIVE LOGITS

ated

0.92

ysis

0.85

esty

0.84

num

0.83

ating

0.83

tion

0.82

ation

0.80

atives

0.76

ayne

0.73

ate

0.72

Activations Density 0.012%

phrases related to problem-solving or fixing issues

instances of the word "fix" and its variations

No Comments

No Known Activations