INDEX

Explanations

In this case, the neuron appears to be looking for words related to criticism or negativity

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

agher

-1.16

reon

-1.01

ulhu

-1.00

orate

-0.98

 Ancients

-0.95

arians

-0.95

inosaur

-0.95

ICAN

-0.94

orians

-0.94

 Mant

-0.94

POSITIVE LOGITS

ball

1.58

ening

1.47

hearted

1.41

ener

1.33

grass

1.32

eners

1.25

cover

1.25

 palate

1.23

heart

1.21

balls

1.16

Activations Density 0.873%

In this case, the neuron appears to be looking for words related to criticism or negativity

No Comments

No Known Activations