INDEX

Explanations

descriptions of injuries and incidents involving hospitals

oai_token-act-pair · gpt-3.5-turbo

keywords related to incidents involving injuries and medical responses

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

DragonMagazine

-0.85

 billions

-0.75

 trillions

-0.74

 ARTICLE

-0.68

 Millions

-0.67

 hypocritical

-0.67

 hither

-0.66

 Thousands

-0.64

 famine

-0.64

 Tribunal

-0.63

POSITIVE LOGITS

nsics

0.90

 remains

0.90

hunt

0.89

EMS

0.89

 hospitalized

0.87

 identified

0.86

 bystanders

0.84

 paramedics

0.81

 remained

0.81

 injuries

0.80

Activations Density 0.350%

descriptions of injuries and incidents involving hospitals

keywords related to incidents involving injuries and medical responses

No Comments

No Known Activations