INDEX

Explanations

phrases related to legislation, political debates, and governmental sessions

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

 marketed

-0.37

pim

-0.37

obyl

-0.35

 maternal

-0.33

 unborn

-0.33

 terr

-0.32

 Shinji

-0.32

 Fans

-0.32

laim

-0.31

 traffickers

-0.30

POSITIVE LOGITS

 adjourn

0.46

member

0.45

ansas

0.41

adelphia

0.40

 recess

0.40

rooms

0.40

woman

0.40

office

0.39

room

0.39

mate

0.38

Activations Density 8.922%

phrases related to legislation, political debates, and governmental sessions

No Comments

No Known Activations