INDEX

Explanations

mentions or references to the drug "paradox."

oai_token-act-pair · gpt-3.5-turbo

references to the opioid antagonist naloxone

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 rece

-0.73

maj

-0.68

CHO

-0.66

 certific

-0.66

 relativity

-0.65

 victimized

-0.65

 disemb

-0.64

 immersion

-0.63

 obligated

-0.63

 fries

-0.62

POSITIVE LOGITS

imity

1.18

eus

1.01

es

0.99

ox

0.92

idation

0.91

edo

0.91

esville

0.90

amines

0.90

imus

0.90

nard

0.90

Activations Density 0.011%

mentions or references to the drug "paradox."

references to the opioid antagonist naloxone

No Comments

No Known Activations