INDEX

Explanations

contractions of "should not"

oai_token-act-pair · gpt-3.5-turbo

phrases expressing prohibition or things that should not occur

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

croft

-0.70

lip

-0.67

dress

-0.65

chel

-0.64

 Puzzle

-0.63

ULT

-0.62

 Mutant

-0.61

 Carbuncle

-0.57

 Chains

-0.57

OTOS

-0.56

POSITIVE LOGITS

't

1.18

ned

0.85

 rightfully

0.83

ny

0.79

ighed

0.76

0.75

tyard

0.73

\'

0.71

Override

0.69

ublic

0.69

Activations Density 0.006%

contractions of "should not"

phrases expressing prohibition or things that should not occur

No Comments

No Known Activations