|
gptkbp:instance_of
|
gptkb:AI_technology
|
|
gptkbp:aiadvantage
|
strategic decision making
|
|
gptkbp:aibehavioral_adaptation
|
real-time adjustments
|
|
gptkbp:aichallenge
|
human intelligence
|
|
gptkbp:aidevelopment
|
pioneering in esports
|
|
gptkbp:aienhancements
|
adaptive
competitive level
|
|
gptkbp:aigame_learning
|
through gameplay experience
|
|
gptkbp:aigame_play_style
|
aggressive and strategic
|
|
gptkbp:aiimpact_on
|
future of esports
|
|
gptkbp:aiinteraction
|
with human players
|
|
gptkbp:aiplayer_adaptability
|
highly adaptable to strategies
|
|
gptkbp:aiplayer_behavior
|
cooperative and competitive
|
|
gptkbp:aiplayer_count
|
five AI players
|
|
gptkbp:aiplayer_development
|
continuous improvement
|
|
gptkbp:aiplayer_skill
|
highly skilled
|
|
gptkbp:airesearch_contribution
|
advancements in AI capabilities
|
|
gptkbp:airesearch_field
|
machine learning in gaming
|
|
gptkbp:airesearch_focus
|
game strategy optimization
|
|
gptkbp:airesearch_outcome
|
new insights into AI capabilities.
|
|
gptkbp:aitraining_challenge
|
complex decision making
|
|
gptkbp:aitraining_duration
|
several months
|
|
gptkbp:aitraining_method
|
reinforcement learning with human feedback
|
|
gptkbp:aitraining_methodology
|
iterative training process
|
|
gptkbp:aitraining_success
|
achieved high win rates against humans
|
|
gptkbp:audience
|
mixed reviews
|
|
gptkbp:collaboration
|
with professional Dota 2 players
|
|
gptkbp:community_feedback
|
enthusiastic
|
|
gptkbp:community_impact
|
inspired AI research in gaming
|
|
gptkbp:developed_by
|
gptkb:Open_AI
|
|
gptkbp:event_participation
|
Dota 2 tournaments
|
|
gptkbp:first_match
|
against Team OG
|
|
gptkbp:future_prospects
|
AI in other games
|
|
gptkbp:game_community
|
Dota 2 community
|
|
gptkbp:game_engine
|
gptkb:Source_engine
|
|
gptkbp:game_length
|
approximately 30-60 minutes
|
|
gptkbp:game_type
|
gptkb:arena
|
|
gptkbp:goal
|
compete with human players
|
|
gptkbp:has_role
|
carry, support, offlaner, midlaner, jungler
|
|
https://www.w3.org/2000/01/rdf-schema#label
|
Open AI Dota 2
|
|
gptkbp:influence_on
|
AI development strategies
|
|
gptkbp:intelligence
|
gptkb:Artificial_Intelligence
|
|
gptkbp:is_a_supporter_of
|
gptkb:2019_International_Dota_2_Championship
|
|
gptkbp:map
|
Dota 2 map
|
|
gptkbp:match_type
|
best of three
|
|
gptkbp:mission
|
destroy enemy towers and ancient
|
|
gptkbp:notable_achievement
|
defeated professional players
|
|
gptkbp:notable_match
|
against Evil Geniuses
|
|
gptkbp:objective
|
team-based strategy
|
|
gptkbp:performance
|
win rate
|
|
gptkbp:provides_information_on
|
Dota 2 gameplay data
|
|
gptkbp:related_to
|
gptkb:Dota_2
|
|
gptkbp:release_year
|
gptkb:2018
|
|
gptkbp:research_areas
|
AI in competitive gaming
|
|
gptkbp:skills
|
high-level play
|
|
gptkbp:strategic_element
|
team coordination
|
|
gptkbp:strategy
|
dynamic and evolving
|
|
gptkbp:team
|
five-player team
|
|
gptkbp:training
|
self-play
|
|
gptkbp:training_programs
|
simulated Dota 2 matches
|
|
gptkbp:uses_technology
|
reinforcement learning
|
|
gptkbp:bfsParent
|
gptkb:gymnasium
|
|
gptkbp:bfsLayer
|
4
|