INDEX

Explanations

references to government-issued instructions or commands, particularly executive orders

oai_token-act-pair · gpt-3.5-turbo

instances of executive orders and related terminology

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

 Argon

-0.78

dden

-0.71

apest

-0.69

espie

-0.68

olls

-0.68

 Cats

-0.67

rown

-0.67

Unity

-0.67

stocks

-0.66

rum

-0.66

POSITIVE LOGITS

 prohibiting

1.04

 authorizing

1.01

 issued

1.00

 granting

0.96

 restricting

0.95

 barring

0.91

 banning

0.91

 decree

0.90

 prohibits

0.88

 restraining

0.87

Activations Density 0.051%

references to government-issued instructions or commands, particularly executive orders

instances of executive orders and related terminology

No Comments

No Known Activations