INDEX

Explanations

references to fires

oai_token-act-pair · gpt-3.5-turbo

instances of the word "fires" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

berman

-0.68

eson

-0.67

IQ

-0.65

mx

-0.64

IU

-0.64

ļéĨĴ

-0.63

ographical

-0.60

ophone

-0.59

ppy

-0.59

atri

-0.59

POSITIVE LOGITS

 fires

4.00

 Fires

2.73

fires

2.63

 fire

2.30

 wildfires

2.26

 flames

1.93

fire

1.87

Fire

1.67

 blaze

1.62

 fired

1.59

Activations Density 0.014%

references to fires

instances of the word "fires" in various contexts

No Comments

No Known Activations