[Theory] REMINDER: 1/13 Talks at TTIC: Kush Bhatia, UC Berkeley

[Mary Marre]

*When:*      Thursday, January 13th at* 11:00 am CT*



*Where:*     Zoom Virtual Talk (*register in advance here
<https://uchicagogroup.zoom.us/webinar/register/WN_EBg_DaYoTWycE8aDhN6e_A>*)


*Who: *       Kush Bhatia, UC Berkeley



*Title:*        Learning When Objectives Are Hard to Specify


*Abstract: *

Real world deployment of learning systems which interact with humans
requires aligning what these systems optimize for with underlying human
objectives and values. A major hurdle towards accomplishing this has been
that it is hard for humans to precisely specify what it means to do the
desired task well.

In the first part, we will take a multi-criteria viewpoint towards these
underlying objectives and develop a framework for selection of such
value-aligned models when the data comprises pairwise comparisons across
multiple different criteria. In the second part, I will present our work on
understanding the consequence of over optimizing a misspecified objective.
We find evidence about the existence of phase transitions which could pose
challenges to safe deployment of such learning systems.



*Host*: *Nati Srebro* <nati at ttic.edu>



Mary C. Marre
Faculty Administrative Support
*Toyota Technological Institute*
*6045 S. Kenwood Avenue*
*Chicago, IL  60637*
*mmarre at ttic.edu <mmarre at ttic.edu>*


On Fri, Jan 7, 2022 at 10:53 AM Mary Marre <mmarre at ttic.edu> wrote:

> *When:*      Thursday, January 13th at* 11:00 am CT*
>
>
>
> *Where:*     Zoom Virtual Talk (*register in advance here
> <https://uchicagogroup.zoom.us/webinar/register/WN_EBg_DaYoTWycE8aDhN6e_A>*
> )
>
>
> *Who: *       Kush Bhatia, UC Berkeley
>
>
>
> *Title:*        Learning When Objectives Are Hard to Specify
>
>
> *Abstract: *
>
> Real world deployment of learning systems which interact with humans
> requires aligning what these systems optimize for with underlying human
> objectives and values. A major hurdle towards accomplishing this has been
> that it is hard for humans to precisely specify what it means to do the
> desired task well.
>
> In the first part, we will take a multi-criteria viewpoint towards these
> underlying objectives and develop a framework for selection of such
> value-aligned models when the data comprises pairwise comparisons across
> multiple different criteria. In the second part, I will present our work on
> understanding the consequence of over optimizing a misspecified objective.
> We find evidence about the existence of phase transitions which could pose
> challenges to safe deployment of such learning systems.
>
>
>
> *Host*: *Nati Srebro* <nati at ttic.edu>
>
>
>
> Mary C. Marre
> Faculty Administrative Support
> *Toyota Technological Institute*
> *6045 S. Kenwood Avenue*
> *Chicago, IL  60637*
> *mmarre at ttic.edu <mmarre at ttic.edu>*
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.cs.uchicago.edu/pipermail/theory/attachments/20220112/34136b85/attachment.html>