INDEX

Explanations

references to tubes

oai_token-act-pair · gpt-3.5-turbo

references to various types of tubes

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

earable

-0.70

 Lowry

-0.69

eting

-0.68

nee

-0.68

vironment

-0.67

 IEEE

-0.66

cific

-0.66

Vu

-0.64

common

-0.63

ificantly

-0.63

POSITIVE LOGITS

 tube

1.21

 tubes

1.08

tube

0.99

 Tube

0.96

ength

0.83

 tubing

0.80

Tube

0.80

odon

0.79

ulla

0.77

ulously

0.73

Activations Density 0.005%

references to tubes

references to various types of tubes

No Comments

No Known Activations