INDEX
Explanations
occurrences of laughter and related expressions in conversational contexts
New Auto-Interp
Negative Logits
phet
-0.15
enstein
-0.14
anos
-0.14
ipt
-0.14
onomy
-0.14
.↵↵↵↵
-0.14
enk
-0.14
]+)/
-0.13
obj
-0.13
.↵↵↵↵
-0.13
POSITIVE LOGITS
)
0.34
:)
0.32
]
0.26
}
0.23
")
0.23
ा)
0.22
)
0.22
à¥Ģ)
0.20
_)
0.20
!)
0.20
Activations Density 0.196%