INDEX

Explanations

medical and scientific terms related to research findings and associations

oai_token-act-pair · gpt-3.5-turbo

phrases related to scientific findings and associations in research studies

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

thora

-0.79

ighting

-0.73

Peb

-0.71

arta

-0.70

 Horizons

-0.68

bish

-0.67

 IMAGES

-0.67

 Sloan

-0.67

haven

-0.66

 Afgh

-0.66

POSITIVE LOGITS

 correlated

1.48

 associated

1.43

 predictive

1.42

 positively

1.35

 statistically

1.20

 negatively

1.19

 predictor

1.18

 strongly

1.14

 protective

1.13

 significantly

1.13

Activations Density 0.139%

medical and scientific terms related to research findings and associations

phrases related to scientific findings and associations in research studies

No Comments

No Known Activations