INDEX

Explanations

mentions of different types of meats, particularly pork

oai_token-act-pair · gpt-3.5-turbo

mentions of pork

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

DCS

-0.78

EMBER

-0.76

 Stanton

-0.76

 Insp

-0.73

 Lauder

-0.72

 Occupations

-0.71

orsi

-0.71

idges

-0.69

nces

-0.68

 Dresden

-0.67

POSITIVE LOGITS

bean

1.04

 belly

1.04

meat

1.01

 chops

0.89

 chop

0.86

 roast

0.85

 barrel

0.83

sty

0.82

 pork

0.81

stra

0.79

Activations Density 0.007%

mentions of different types of meats, particularly pork

mentions of pork

No Comments

No Known Activations