INDEX

Explanations

number pairs or sets

oai_token-act-pair · gpt-3.5-turbo

phrases referring to pairs or groups of items

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

racuse

-0.84

fw

-0.75

Downloadha

-0.72

Fu

-0.70

np

-0.67

Ã©r

-0.65

akin

-0.64

biz

-0.63

obbies

-0.63

abus

-0.62

POSITIVE LOGITS

 extremes

1.01

 paragraphs

0.85

 alone

0.83

 gentlemen

0.82

 separately

0.82

 factors

0.81

 halves

0.80

 endeavors

0.80

 individually

0.79

worldly

0.76

Activations Density 0.069%

number pairs or sets

phrases referring to pairs or groups of items

No Comments

No Known Activations