INDEX

Explanations

information related to suspects or individuals under investigation

oai_token-act-pair · gpt-3.5-turbo

references to suspects in a criminal context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

issance

-0.86

atters

-0.70

psey

-0.69

cast

-0.69

parable

-0.69

atism

-0.68

galitarian

-0.67

cision

-0.66

 stead

-0.65

equality

-0.64

POSITIVE LOGITS

 suspects

0.87

 arrested

0.80

FBI

0.77

 suspect

0.77

 warrants

0.76

 apprehended

0.76

 confessed

0.74

 mastermind

0.74

 kidn

0.73

 lineup

0.72

Activations Density 0.054%

information related to suspects or individuals under investigation

references to suspects in a criminal context

No Comments

No Known Activations