INDEX

Explanations

words related to building structures, specifically focusing on floors and ceilings

oai_token-act-pair · gpt-3.5-turbo

references to flooring materials or structures

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

eday

-0.73

ega

-0.72

elson

-0.71

 Mehran

-0.68

NULL

-0.68

ague

-0.67

enegger

-0.65

Ã±o

-0.64

 Schwarzenegger

-0.63

ALS

-0.63

POSITIVE LOGITS

boards

1.20

plan

1.05

nuts

1.02

floor

1.00

 mats

1.00

 tiles

0.95

 floor

0.95

pace

0.90

nut

0.90

 Floor

0.90

Activations Density 0.037%

words related to building structures, specifically focusing on floors and ceilings

references to flooring materials or structures

No Comments

No Known Activations