INDEX

Explanations

references to emergency or alarming situations, particularly related to sirens and screeching noises

oai_token-act-pair · gpt-3.5-turbo

words related to emergency signals, particularly sirens

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

berman

-0.79

eneg

-0.75

¥

-0.71

erity

-0.71

ĵĺ

-0.68

 untreated

-0.67

absolute

-0.64

±

-0.62

lege

-0.62

rival

-0.62

POSITIVE LOGITS

iren

1.27

irens

1.17

 warnings

0.90

unci

0.86

wip

0.76

crow

0.76

cia

0.75

 caller

0.75

 Warn

0.74

 Krug

0.72

Activations Density 0.006%

references to emergency or alarming situations, particularly related to sirens and screeching noises

words related to emergency signals, particularly sirens

No Comments

No Known Activations