Robocrunch
Sergey Levine
@svlevine
11/24/2021
Value
function
spaces
(VFS)
uses
lowlevel
primitives
to
form
a
state
representation
in
terms
of
their
"affordances"

the
value
functions
of
the
primitives
serve
as
the
state.
This
turns
out
to
really
improve
generalization
in
hierarchical
RL!
https://t.co/yJqJwwCT6r
Short
ðŸ§µ>
25
154
Related
Elias Najarro
@enasmel
6 hours
"Tailoring:
encoding
inductive
biases
by
optimising
unsupervised
objectives
at
prediction
time"
by
@FerranAlet
Gist:
Add
extra
loss
function
during
inference.
New
loss
"tailors"
the
NN
parameters
to
improve
NN
representation
and
hence
performance.
Paper:
https://t.co/hc678ebSAl
3
Alex Nichol
@unixpickle
11/25/2021
I
loved
learning
math
as
a
kid
because
I
knew
I'd
be
consuming
well
thoughtout
abstractions.
Math
is
full
of
really
useful
primitives
developed
over
thousand
of
years,
and
I
knew
I
was
in
good
hands
if
I
learned
them.
3
AK
@ak92501
11/30/2021
Improving
Zeroshot
Generalization
in
Offline
Reinforcement
Learning
using
Generalized
Similarity
Functions
abs:
https://t.co/5B163MSLHN
15
Demis Hassabis
@demishassabis
12/3/2021
So
Iâ€™m
really
proud
of
our
latest
@Nature
paper
https://t.co/L4UTPV39PE
(also
on
the
cover!)
detailing
our
fantastic
collaborations
with
Maths
Profs
Williamson,
Lackenby,
and
Juhasz
to
make
surprising
and
significant
discoveries
in
topology
and
representation
theory
12
93