INDEX

Explanations

instances of endangerment or potential harm in a scenario

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

 appreciate

-0.39

oult

-0.37

 appreciated

-0.36

uno

-0.36

annis

-0.35

Fram

-0.35

 evenly

-0.35

 Edited

-0.35

winner

-0.34

rikes

-0.34

POSITIVE LOGITS

 endanger

0.54

orage

0.49

ously

0.47

angering

0.46

rance

0.45

 Hazard

0.45

 bystanders

0.45

mental

0.45

ment

0.44

azard

0.43

Activations Density 6.889%

instances of endangerment or potential harm in a scenario

No Comments

No Known Activations