INDEX

Explanations

words related to self-identity and self-description

oai_token-act-pair · gpt-3.5-turbo

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

 Nights

-0.79

 Ashe

-0.77

 XIII

-0.74

 GOODMAN

-0.74

ī

-0.70

 Syndicate

-0.69

IUM

-0.68

 Rouge

-0.67

 Slay

-0.67

LAT

-0.66

POSITIVE LOGITS

destruct

1.30

 destruct

1.00

conscious

1.00

ridges

0.97

same

0.94

 explanatory

0.92

lessly

0.91

upload

0.89

 esteem

0.84

pecially

0.81

Activations Density 0.041%

words related to self-identity and self-description

No Comments

No Known Activations