INDEX

Explanations

words related to human behaviors and experiences

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Configuration

neuronpedia/gpt2-small__res_scefr-ajt/6-res_scefr-ajt

Prompts (Dashboard)

12,288 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

46,080

Data Type

torch.float32

Hook Point

blocks.6.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

apollo-research/Skylion007-openwebtext-tokenizer-gpt2

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

PsyNetMessage

-1.01

inav

-0.98

 fray

-0.81

lake

-0.79

NetMessage

-0.78

 Sochi

-0.78

eday

-0.78

 Lilly

-0.76

Lag

-0.76

orrow

-0.75

POSITIVE LOGITS

 able

1.63

 bitten

1.15

 aware

1.09

 eaten

1.08

 notified

1.05

 unable

1.05

 briefed

1.03

 rewarded

1.01

 immersed

1.01

que

0.99

Activations Density 0.573%

words related to human behaviors and experiences

No Comments

No Known Activations