INDEX

Explanations

words related to physical coverings or disguises

oai_token-act-pair · gpt-3.5-turbo

terms associated with legal and regulatory contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

enne

-0.62

orically

-0.61

ategory

-0.59

Ø¯

-0.59

ibraries

-0.59

olver

-0.58

anamo

-0.56

 DRAG

-0.56

ON

-0.55

Ī

-0.55

POSITIVE LOGITS

 basis

1.28

 occasions

1.20

 behalf

1.11

 occasion

1.00

 doorstep

0.95

 periphery

0.89

 shelf

0.89

 shelves

0.87

 verge

0.86

 pedest

0.85

Activations Density 0.498%

words related to physical coverings or disguises

terms associated with legal and regulatory contexts

No Comments

No Known Activations