INDEX

Explanations

scientific and medical terms, possibly related to research studies and mechanisms

oai_token-act-pair · gpt-3.5-turbo

scientific terminology and complex concepts related to medical studies

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

Kids

-0.74

gee

-0.74

vernment

-0.73

ulhu

-0.72

wives

-0.72

 Scouts

-0.70

Britain

-0.69

foundland

-0.67

illions

-0.67

 Vanity

-0.66

POSITIVE LOGITS

 inhibition

1.34

 modulation

1.22

 oxidative

1.18

 blockade

1.17

cyt

1.16

 endogenous

1.13

 pharmac

1.13

 Î²

1.12

 phosph

1.12

 GABA

1.12

Activations Density 0.261%

scientific and medical terms, possibly related to research studies and mechanisms

scientific terminology and complex concepts related to medical studies

No Comments

No Known Activations