BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//pretalx//pretalx.devconf.info//devconf-us-2025//talk//KQJVGU
BEGIN:VTIMEZONE
TZID:EST
BEGIN:STANDARD
DTSTART:20001029T030000
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=10;UNTIL=20061029T070000Z
TZNAME:EST
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
END:STANDARD
BEGIN:STANDARD
DTSTART:20071104T030000
RRULE:FREQ=YEARLY;BYDAY=1SU;BYMONTH=11
TZNAME:EST
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
END:STANDARD
BEGIN:DAYLIGHT
DTSTART:20000402T030000
RRULE:FREQ=YEARLY;BYDAY=1SU;BYMONTH=4;UNTIL=20060402T080000Z
TZNAME:EDT
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
END:DAYLIGHT
BEGIN:DAYLIGHT
DTSTART:20070311T030000
RRULE:FREQ=YEARLY;BYDAY=2SU;BYMONTH=3
TZNAME:EDT
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
END:DAYLIGHT
END:VTIMEZONE
BEGIN:VEVENT
UID:pretalx-devconf-us-2025-KQJVGU@pretalx.devconf.info
DTSTART;TZID=EST:20250920T100000
DTEND;TZID=EST:20250920T101500
DESCRIPTION:This presentation examines how reinforcement fine-tuning (RFT) 
 can be applied to enhance agentic reasoning in large language models. We b
 egin by motivating the need for explicit reasoning\, highlighting its role
  in reliability\, transparency\, and generalization across complex tasks. 
 Building on this foundation\, we showcase an experimental case study using
  the Wordle game as a testbed\, demonstrating how RFT guides models to dev
 elop structured reasoning strategies rather than relying on shallow patter
 n matching. Finally\, we present a concise survey of related approaches\, 
 situating RFT within the broader landscape of reinforcement learning for r
 easoning\, and outlining open directions for improving robustness and effi
 ciency in agentic systems.
DTSTAMP:20260311T010341Z
LOCATION:Ladd Room (Capacity 170)
SUMMARY:Agentic Reasonings with Reinforcement Learning - Andy Xie
URL:https://pretalx.devconf.info/devconf-us-2025/talk/KQJVGU/
END:VEVENT
END:VCALENDAR