INDEX

Explanations

specific keywords related to projects, legislation, and product features

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Configuration

neuronpedia/gpt2-small__res_scl-ajt/6-res_scl-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

 playthrough

-0.19

guard

-0.18

isc

-0.17

ko

-0.17

 corners

-0.17

 swearing

-0.17

ann

-0.16

due

-0.16

 vide

-0.16

href

-0.16

POSITIVE LOGITS

ILE

0.31

IONS

0.30

IVES

0.30

ION

0.28

URE

0.28

ERY

0.27

URES

0.26

IVE

0.26

NAME

0.25

ATIONS

0.25

Activations Density 12.850%

specific keywords related to projects, legislation, and product features

No Comments

No Known Activations