INDEX

Explanations

import, castings, PLA

This neuron fires strongly on morphemes and fragments of chemical or biochemical compound names.

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 Privacidade

0.19

खरी

0.18

chrane

0.18

responses

0.17

)`

0.17

↵↵

0.17

("|

0.17

successor

0.16

 veuillez

0.16

oruč

0.16

POSITIVE LOGITS

这个

0.25

 meiner

0.22

 της

0.22

ק

0.22

ların

0.22

ка

0.21

這個

0.21

rijke

0.21

一個

0.21

นี้

0.20

Activations Density 2.560%