INDEX

Explanations

legal and regulatory terms or statements

oai_token-act-pair · gpt-3.5-turbo

legal terminology and phrases associated with regulations and laws

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

oom

-0.73

ovie

-0.72

oice

-0.71

zinski

-0.70

ortun

-0.69

cade

-0.67

enes

-0.66

MK

-0.65

Sit

-0.64

icipated

-0.64

POSITIVE LOGITS

 governs

1.06

 rules

1.03

 regulating

1.02

 governing

1.01

 regulations

1.00

 prohibiting

1.00

enforcement

0.97

 applicable

0.95

 promulg

0.94

 prohibitions

0.93

Activations Density 0.790%

legal and regulatory terms or statements

legal terminology and phrases associated with regulations and laws

No Comments

No Known Activations