INDEX

Explanations

verbs related to actions aimed at avoiding or stopping something

oai_token-act-pair · gpt-3.5-turbo

phrases related to prevention

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

ammy

-0.87

geist

-0.75

edded

-0.66

aturday

-0.63

issance

-0.62

 Colleges

-0.62

 Pricing

-0.61

subject

-0.61

 Snow

-0.61

 Paid

-0.60

POSITIVE LOGITS

 inhib

0.89

ative

0.86

 prevent

0.85

 allev

0.81

 preventing

0.80

 detection

0.79

 regress

0.78

 prevented

0.74

essim

0.73

 duplicate

0.73

Activations Density 0.014%

verbs related to actions aimed at avoiding or stopping something

phrases related to prevention

No Comments

No Known Activations