INDEX

Explanations

instances where people are described or mentioned as wearing specific attire or accessories

oai_token-act-pair · gpt-3.5-turbo

references to the act of wearing clothing

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

rates

-0.70

lishes

-0.68

Â¢

-0.65

arta

-0.65

phabet

-0.64

Region

-0.63

oi

-0.62

ago

-0.61

 Macro

-0.60

addy

-0.60

POSITIVE LOGITS

 wearing

3.58

 wear

2.16

 wore

2.15

 wears

2.09

 dressed

1.82

 worn

1.77

 wielding

1.69

 clad

1.68

 wearer

1.64

 Wear

1.61

Activations Density 0.012%

instances where people are described or mentioned as wearing specific attire or accessories

references to the act of wearing clothing

No Comments

No Known Activations