INDEX

Explanations

keywords related to dangers or risks

oai_token-act-pair · gpt-3.5-turbo

references to danger or risk

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

ergy

-0.82

eenth

-0.76

Ķ

-0.73

char

-0.71

slice

-0.69

olitan

-0.69

ets

-0.69

rix

-0.69

edited

-0.69

ann

-0.68

POSITIVE LOGITS

 danger

1.20

 Danger

1.16

 endanger

1.08

 menace

0.97

 mosqu

0.91

 undermin

0.91

 peril

0.90

 dangers

0.87

danger

0.87

 jeopardy

0.83

Activations Density 0.006%

keywords related to dangers or risks

references to danger or risk

No Comments

No Known Activations