INDEX

Explanations

mentions of the drug "cocaine."

oai_token-act-pair · gpt-3.5-turbo

mentions of cocaine and its related context

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

igating

-0.78

hof

-0.76

lying

-0.73

 Arbor

-0.73

soType

-0.71

laus

-0.71

ĪĴ

-0.70

aer

-0.70

Bird

-0.70

 Worcester

-0.68

POSITIVE LOGITS

 cocaine

1.15

 residue

1.09

 addiction

0.96

 addicts

0.95

 overdose

0.94

stasy

0.91

 addict

0.91

 overdoses

0.89

 sulf

0.88

 traffickers

0.85

Activations Density 0.011%

mentions of the drug "cocaine."

mentions of cocaine and its related context

No Comments

No Known Activations