INDEX

Explanations

words related to alarms or alerts

oai_token-act-pair · gpt-3.5-turbo

references to alarms or alert mechanisms

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

arnaev

-0.73

lease

-0.71

buff

-0.69

merga

-0.68

lections

-0.68

hews

-0.68

isans

-0.67

employment

-0.66

tein

-0.64

elson

-0.64

POSITIVE LOGITS

 bells

1.28

ingly

1.18

 clock

1.02

 Clock

0.99

 alarm

0.98

 clocks

0.98

clock

0.96

 alarms

0.95

Clock

0.94

ously

0.86

Activations Density 0.029%

words related to alarms or alerts

references to alarms or alert mechanisms

No Comments

No Known Activations