INDEX

Explanations

words related to terrorism and violent attacks

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

ãģ¦

-0.56

 needles

-0.48

serving

-0.46

 ginger

-0.45

 chloride

-0.43

 plastics

-0.43

manship

-0.42

VERTISEMENT

-0.42

 sterling

-0.42

 prest

-0.41

POSITIVE LOGITS

abad

0.75

urous

0.62

ists

0.60

rieg

0.59

raq

0.58

bsp

0.58

ology

0.58

opsy

0.56

ãĥĦ

0.56

ist

0.56

Activations Density 7.921%

words related to terrorism and violent attacks

No Comments

No Known Activations