INDEX

Explanations

words related to air travel incidents, passengers, crew, and injuries

oai_token-act-pair · gpt-3.5-turbo

references to people involved in an incident or event

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

Ĥİ

-0.74

ukong

-0.72

LOS

-0.71

¥µ

-0.71

orously

-0.68

chapter

-0.67

inventoryQuantity

-0.67

é¾įå¥ĳå£«

-0.66

limits

-0.66

 Effective

-0.64

POSITIVE LOGITS

 paramed

0.93

 crew

0.88

 drummer

0.86

 camer

0.85

 paramedics

0.84

 bystanders

0.83

 parach

0.82

 passengers

0.82

 juveniles

0.81

 bystand

0.77

Activations Density 0.463%

words related to air travel incidents, passengers, crew, and injuries

references to people involved in an incident or event

No Comments

No Known Activations