INDEX

Explanations

phrases related to honorable mentions and honorable conduct

oai_token-act-pair · gpt-3.5-turbo

terms associated with legal or disorderly conduct

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

oise

-0.95

efully

-0.95

emouth

-0.86

caster

-0.85

isu

-0.81

oir

-0.81

rays

-0.80

amar

-0.80

emi

-0.78

iago

-0.77

POSITIVE LOGITS

 disorderly

0.93

backer

0.84

lihood

0.82

 affairs

0.74

 conduct

0.73

LO

0.70

 asses

0.69

 actions

0.69

 handc

0.68

EMENT

0.65

Activations Density 0.034%

phrases related to honorable mentions and honorable conduct

terms associated with legal or disorderly conduct

No Comments

No Known Activations