INDEX

Explanations

uncommon or unusual occurrences or entities described as "strange"

oai_token-act-pair · gpt-3.5-turbo

descriptors related to unusual or bizarre elements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

($

-0.78

 outweigh

-0.72

payer

-0.69

aper

-0.68

 provides

-0.68

Â®

-0.67

 Prep

-0.66

 Priority

-0.66

buster

-0.65

âĢ

-0.65

POSITIVE LOGITS

 strange

3.11

 weird

2.23

 bizarre

2.10

odd

2.05

 mysterious

2.01

 peculiar

1.99

 strangely

1.88

Strange

1.81

 unusual

1.80

 inexplicable

1.79

Activations Density 0.025%

uncommon or unusual occurrences or entities described as "strange"

descriptors related to unusual or bizarre elements

No Comments

No Known Activations