INDEX

Explanations

mentions of the television program "60 Minutes"

oai_token-act-pair · gpt-3.5-turbo

references to the television program "60 Minutes."

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

 inherit

-0.71

 discrimination

-0.68

 breed

-0.65

rav

-0.64

 bridge

-0.64

 explo

-0.64

 root

-0.64

 adoptive

-0.63

 sword

-0.63

 inher

-0.63

POSITIVE LOGITS

 Minutes

4.30

 Hours

2.00

 Seconds

1.96

 minutes

1.54

 Minute

1.37

 Moments

1.32

Hours

1.21

 Days

1.21

 Months

1.19

seconds

1.18

Activations Density 0.015%

mentions of the television program "60 Minutes"

references to the television program "60 Minutes."

No Comments

No Known Activations