INDEX

Explanations

phrases related to physical and medical conditions, such as "paralyzed" or "epileptic"

oai_token-act-pair · gpt-3.5-turbo

terms related to physical disability and recovery

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

arella

-0.65

 Kelvin

-0.64

ipop

-0.62

 Dragonbound

-0.62

 Hamm

-0.60

 adam

-0.58

heartedly

-0.58

Ou

-0.57

 Danish

-0.57

 anecd

-0.57

POSITIVE LOGITS

ments

1.55

ment

1.49

mented

1.42

mental

1.38

ller

1.34

llers

1.30

mentation

1.23

vous

1.20

llor

1.18

ptive

1.09

Activations Density 0.114%

phrases related to physical and medical conditions, such as "paralyzed" or "epileptic"

terms related to physical disability and recovery

No Comments

No Known Activations