INDEX

Explanations

words related to the software program "Word."

oai_token-act-pair · gpt-3.5-turbo

references to Microsoft Word

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

olson

-0.81

 unnecess

-0.81

\/\/

-0.80

ocal

-0.77

Interstitial

-0.75

ORK

-0.74

olulu

-0.73

âĸ

-0.71

OUR

-0.70

etheus

-0.66

POSITIVE LOGITS

 Word

1.35

Word

1.16

sworth

1.14

press

1.13

Pad

0.82

word

0.81

Press

0.81

Spread

0.76

Perfect

0.74

Rank

0.73

Activations Density 0.004%

words related to the software program "Word."

references to Microsoft Word

No Comments

No Known Activations