INDEX

Explanations

reinforcement learning environments

The neuron fires on mentions of reinforcement-learning environment names and related library identifiers (e.g. “gym,” “Box2D,” “BipedalWalker,” “env”).

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 solche

-1.40

 solchen

-1.38

 Süden

-1.37

 soziale

-1.29

 необходимые

-1.27

 geisti

-1.25

 wichtigen

-1.24

June

-1.24

 wichtige

-1.23

os

-1.20

POSITIVE LOGITS

1.46

]+

1.42

[:

1.37

ⅼ

1.37

</sup>

1.35

('/')

1.33

terdam

1.30

😐

1.30

สง

1.28

 Viitattu

1.26

Activations Density 0.016%