INDEX

Explanations

phrases related to approval or authorization

oai_token-act-pair · gpt-3.5-turbo

phrases related to approvals, sign-offs, or endorsements in governmental or official contexts

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

 simmer

-0.68

 concentrating

-0.66

 submer

-0.65

gypt

-0.64

 stagnant

-0.64

stery

-0.63

 Stir

-0.63

 pests

-0.63

pell

-0.62

 cooking

-0.62

POSITIVE LOGITS

reed

0.81

*/(

0.80

à¼

0.79

ellation

0.78

 waiver

0.75

 Cosponsors

0.75

acts

0.73

oak

0.72

binding

0.72

 veto

0.70

Activations Density 0.110%

phrases related to approval or authorization

phrases related to approvals, sign-offs, or endorsements in governmental or official contexts

No Comments

No Known Activations