INDEX

Explanations

measurements of weight or quantity, particularly in tons

oai_token-act-pair · gpt-3.5-turbo

references to weight measurements, specifically in tons

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

Journal

-0.74

cknow

-0.73

 recol

-0.72

Scotland

-0.71

stood

-0.70

bro

-0.69

leader

-0.68

eer

-0.66

gotten

-0.64

 Frontier

-0.63

POSITIVE LOGITS

 tons

1.16

umen

0.85

 tract

0.82

neau

0.81

ension

0.80

ongs

0.80

urgical

0.78

 tires

0.77

ensical

0.76

umes

0.76

Activations Density 0.004%

measurements of weight or quantity, particularly in tons

references to weight measurements, specifically in tons

No Comments

No Known Activations