INDEX

Explanations

numeric values associated with weight units

oai_token-act-pair · gpt-3.5-turbo

phrases related to weight measurements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

 Occupations

-0.80

mberg

-0.69

éĹ

-0.64

employment

-0.63

 Acad

-0.63

 Nasa

-0.60

 partName

-0.60

IDES

-0.59

arta

-0.59

actionDate

-0.59

POSITIVE LOGITS

pound

1.28

 pound

1.02

 sterling

0.98

 pounds

0.93

 Pound

0.93

lbs

0.88

cake

0.85

lb

0.84

tackle

0.82

pole

0.82

Activations Density 0.005%

numeric values associated with weight units

phrases related to weight measurements

No Comments

No Known Activations