INDEX

Explanations

words related to electronic sensors

oai_token-act-pair · gpt-3.5-turbo

references to sensors and their applications

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

paragraph

-0.74

eve

-0.73

 Bald

-0.71

enburg

-0.71

tein

-0.67

press

-0.67

 Virgin

-0.66

swick

-0.66

 Gazette

-0.65

 Afghanistan

-0.65

POSITIVE LOGITS

 sensors

1.15

ionics

0.95

 microphones

0.94

 enthusi

0.92

 sensor

0.88

wip

0.85

ynthesis

0.81

 detects

0.79

 aperture

0.79

 reluct

0.79

Activations Density 0.010%

words related to electronic sensors

references to sensors and their applications

No Comments

No Known Activations