INDEX

Explanations

references to specific laws or legal regulations

oai_token-act-pair · gpt-3.5-turbo

references to laws and regulations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

yss

-0.84

Naz

-0.68

arthed

-0.67

tsky

-0.67

atha

-0.67

joy

-0.65

ortun

-0.64

 Mous

-0.63

Mul

-0.62

alin

-0.62

POSITIVE LOGITS

 enforcement

1.27

 prohibiting

1.01

 governing

0.99

enforcement

0.99

 Enforcement

0.98

 violations

0.96

 regulating

0.95

 enacted

0.93

 laws

0.89

 abiding

0.89

Activations Density 0.024%

references to specific laws or legal regulations

references to laws and regulations

No Comments

No Known Activations