INDEX

Explanations

references to widespread health issues or crises, particularly epidemics

oai_token-act-pair · gpt-3.5-turbo

terms related to outbreaks or widespread issues

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

nda

-0.83

partial

-0.76

earchers

-0.73

akov

-0.72

ighters

-0.72

boa

-0.71

ancies

-0.71

gem

-0.69

prints

-0.68

cius

-0.67

POSITIVE LOGITS

 epidemic

1.32

 Outbreak

0.97

 outbreak

0.96

 scourge

0.90

 endemic

0.84

 epidem

0.80

 plag

0.79

 Ebola

0.74

 virus

0.74

 Virus

0.73

Activations Density 0.009%

references to widespread health issues or crises, particularly epidemics

terms related to outbreaks or widespread issues

No Comments

No Known Activations