INDEX

Explanations

mentions of CIA, FBI, espionage, intelligence, and government activities

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

 Mulcair

-0.61

inances

-0.61

cies

-0.59

igans

-0.59

ths

-0.58

 Bieber

-0.58

own

-0.57

kj

-0.57

kees

-0.55

hots

-0.55

POSITIVE LOGITS

 annex

0.73

 Director

0.67

CIA

0.63

 operatives

0.61

 Headquarters

0.61

CIA

0.60

 covert

0.59

 operative

0.58

 assassinate

0.57

NI

0.57

Activations Density 10.337%

mentions of CIA, FBI, espionage, intelligence, and government activities

No Comments

No Known Activations