INDEX

Explanations

references to legal and political proceedings, including investigations and government actions

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ãĤ¨ãĥ«

-0.69

ãĤ¼

-0.65

Atl

-0.63

Afee

-0.63

GF

-0.62

 girlfriend

-0.62

 intangible

-0.61

jiang

-0.60

 compan

-0.59

Jet

-0.58

POSITIVE LOGITS

 convened

1.04

 unanimously

0.98

 hearings

0.97

 chaired

0.89

 appointed

0.86

member

0.85

 adjourn

0.84

 approving

0.83

 recess

0.81

 deliber

0.81

Activations Density 22.261%

references to legal and political proceedings, including investigations and government actions

No Comments

No Known Activations