INDEX

Explanations

references to different kinds of threats, particularly death threats

oai_token-act-pair · gpt-3.5-turbo

instances of the word "threats" related to various forms of intimidation or danger

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

çĦ

-0.83

arist

-0.81

UX

-0.77

cise

-0.76

bred

-0.73

puted

-0.73

æ©Ł

-0.70

tiny

-0.70

NAS

-0.70

CSS

-0.69

POSITIVE LOGITS

 threats

0.87

 threatening

0.83

 posed

0.79

 against

0.78

 retaliation

0.78

 repr

0.77

 intimidation

0.75

 threatened

0.74

 leveled

0.72

 hotline

0.72

Activations Density 0.030%

references to different kinds of threats, particularly death threats

instances of the word "threats" related to various forms of intimidation or danger

No Comments

No Known Activations