INDEX

Explanations

references to measurements involving height and weight

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 swearing

-0.65

 Goodman

-0.63

 Cutter

-0.61

 prominence

-0.60

brance

-0.59

 descending

-0.59

idence

-0.58

 embarrassment

-0.58

 unpre

-0.58

 Lauder

-0.58

POSITIVE LOGITS

1.39

1.37

1.35

1.31

1.30

1.29

1.28

Activations Density 0.188%

references to measurements involving height and weight

No Comments

No Known Activations