INDEX

Explanations

health-related terms or medical conditions, particularly focusing on chronic conditions

oai_token-act-pair · gpt-3.5-turbo

references to chronic health conditions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

*/(

-0.90

hammad

-0.85

ramid

-0.83

ansk

-0.80

alsa

-0.73

abad

-0.70

Ĭ±

-0.70

dash

-0.68

atography

-0.68

arks

-0.67

POSITIVE LOGITS

 chronic

1.06

 Chronic

0.95

 inflammation

0.79

ãĤ©

0.77

pain

0.70

 homelessness

0.70

 traumatic

0.70

ãĥ¯ãĥ³

0.69

 exacerb

0.69

ãĤ¢ãĥ«

0.68

Activations Density 0.005%

health-related terms or medical conditions, particularly focusing on chronic conditions

references to chronic health conditions

No Comments

No Known Activations