INDEX

Explanations

phrases related to reaching a critical point or threshold

oai_token-act-pair · gpt-3.5-turbo

phrases indicating critical situations or extreme thresholds

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

eston

-0.66

via

-0.62

MN

-0.62

 mailed

-0.62

ocations

-0.62

aughtered

-0.60

aters

-0.59

 Vaughan

-0.58

 Kinnikuman

-0.58

odor

-0.58

POSITIVE LOGITS

manship

1.37

 brink

1.25

terness

1.06

 verge

1.01

wcs

0.82

 bluff

0.80

 perspect

0.79

ï¸

0.73

 edge

0.71

cape

0.70

Activations Density 0.012%

phrases related to reaching a critical point or threshold

phrases indicating critical situations or extreme thresholds

No Comments

No Known Activations