INDEX

Explanations

references to chronological sequences or order

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

bos

-0.83

IPS

-0.82

ISTORY

-0.79

obs

-0.77

mbuds

-0.77

POST

-0.75

restling

-0.75

politics

-0.75

hus

-0.73

oes

-0.73

POSITIVE LOGITS

 installment

1.23

 iteration

1.19

 edition

1.13

 phase

1.09

 incarnation

1.02

 baseman

1.02

 inning

1.02

 consecutive

1.02

 largest

1.01

 anniversary

0.99

Activations Density 4.802%

references to chronological sequences or order

No Comments

No Known Activations