INDEX

Explanations

references to the specific object "Zero"

oai_token-act-pair · gpt-3.5-turbo

references to the Raspberry Pi Zero

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

ells

-0.74

 clerks

-0.72

 cler

-0.72

interstitial

-0.70

fol

-0.70

amn

-0.67

gren

-0.64

 eastern

-0.64

 collagen

-0.64

orn

-0.64

POSITIVE LOGITS

 Zero

4.01

Zero

3.22

zero

2.37

 zero

2.01

Zer

1.20

 Negative

1.15

ERO

1.10

 Eleven

1.09

 Sigma

1.04

 Quantum

1.01

Activations Density 0.013%

references to the specific object "Zero"

references to the Raspberry Pi Zero

No Comments

No Known Activations