INDEX

Explanations

the acronym "OM" preceded by a number

oai_token-act-pair · gpt-3.5-turbo

the abbreviation "OM" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

dil

-0.70

 "$:/

-0.66

 deport

-0.65

 bypass

-0.64

 diluted

-0.63

 lawful

-0.62

hed

-0.62

 Stain

-0.61

 Garland

-0.61

 striker

-0.61

POSITIVE LOGITS

OM

4.14

oms

1.94

om

1.86

OME

1.81

omm

1.51

omb

1.50

OSH

1.41

OH

1.38

oming

1.37

OG

1.37

Activations Density 0.011%

the acronym "OM" preceded by a number

the abbreviation "OM" in various contexts

No Comments

No Known Activations