INDEX

Explanations

terms related to extremist groups and ideologies, including white supremacists, neo-Nazis, and fascist movements

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

anwhile

-0.64

 Rica

-0.59

 Tiff

-0.56

AVG

-0.56

++++++++++++++++

-0.54

ovember

-0.53

pload

-0.52

pport

-0.51

Effective

-0.51

channelAvailability

-0.50

POSITIVE LOGITS

fascist

0.74

colonial

0.72

liberal

0.71

social

0.67

present

0.67

Nazi

0.67

natal

0.66

aligned

0.64

capitalist

0.64

Georg

0.63

Activations Density 8.266%

terms related to extremist groups and ideologies, including white supremacists, neo-Nazis, and fascist movements

No Comments

No Known Activations