INDEX

Explanations

references to technologies or innovations, such as processors, computers, and new systems

oai_token-act-pair · gpt-3.5-turbo

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

 Errors

-0.81

 tends

-0.72

 Procedure

-0.72

 complains

-0.72

 beforehand

-0.72

pse

-0.71

 aggrav

-0.69

autions

-0.68

 Issues

-0.67

 evid

-0.67

POSITIVE LOGITS

 seamless

1.11

 someday

1.10

 wearable

1.07

 smarter

1.05

earable

1.04

 autonomous

1.01

 interoper

1.00

IoT

1.00

 futuristic

0.98

 scalable

0.97

Activations Density 0.601%

references to technologies or innovations, such as processors, computers, and new systems

No Comments

No Known Activations