INDEX

Explanations

adjectives related to intensity or enhancement

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

Goo

-0.63

orers

-0.62

CLA

-0.58

zip

-0.56

opers

-0.55

 craw

-0.55

ANGE

-0.55

emp

-0.55

lore

-0.55

CoC

-0.53

POSITIVE LOGITS

by

1.23

by

0.91

BY

0.88

By

0.87

By

0.83

 anew

0.81

igated

0.77

 exponentially

0.74

 aback

0.73

.</

0.72

Activations Density 0.148%

adjectives related to intensity or enhancement

No Comments

No Known Activations