diff --git a/config.yaml b/config.yaml
index 25d3084..7265c8f 100644
--- a/config.yaml
+++ b/config.yaml
@@ -8,7 +8,7 @@ debug:
     headless: False
     stream_wrapper: False
     init_state: victory_road
-    max_steps: 16
+    max_steps: 20480
     log_frequency: 1
     disable_wild_encounters: True
     disable_ai_actions: True
@@ -289,6 +289,7 @@ rewards:
       exploration_plateau: 0.025
       exploration_lobby: 0.035 # for game corner
       a_press: 0.00001
+      explore_warps: 0.03
 
 
 
diff --git a/pokemonred_puffer/environment.py b/pokemonred_puffer/environment.py
index ff4083d..ba3347f 100644
--- a/pokemonred_puffer/environment.py
+++ b/pokemonred_puffer/environment.py
@@ -251,6 +251,7 @@ def register_hooks(self):
             self.setup_disable_wild_encounters()
         self.pyboy.hook_register(None, "AnimateHealingMachine", self.pokecenter_heal_hook, None)
         self.pyboy.hook_register(None, "OverworldLoopLessDelay", self.overworld_loop_hook, None)
+        self.pyboy.hook_register(None, "CheckWarpsNoCollisionLoop", self.update_warps_hook, None)
 
     def setup_disable_wild_encounters(self):
         bank, addr = self.pyboy.symbol_lookup("TryDoWildEncounter.gotWildEncounterType")
@@ -371,6 +372,7 @@ def init_mem(self):
         self.cut_explore_map = np.zeros(GLOBAL_MAP_SHAPE, dtype=np.float32)
         self.seen_map_ids = np.zeros(256)
         self.seen_npcs = {}
+        self.seen_warps = {}
 
         self.cut_coords = {}
         self.cut_tiles = {}
@@ -652,6 +654,7 @@ def step(self, action):
         elif self.step_count % self.log_frequency == 0:
             info = info | self.agent_stats(action)
         self.required_events = required_events
+        print(self.seen_warps)
 
         obs = self._get_obs()
 
@@ -1136,6 +1139,16 @@ def pokecenter_heal_hook(self, *args, **kwargs):
     def overworld_loop_hook(self, *args, **kwargs):
         self.user_control = True
 
+    def update_warps_hook(self, *args, **kwargs):
+        # current map id, destiation map id, warp id
+        key = (
+            self.read_m("wCurMap"),
+            self.read_m("hWarpDestinationMap"),
+            self.read_m("wDestinationWarpID"),
+        )
+        if key[-1] != 0xFF:
+            self.seen_warps[key] = 1
+
     def cut_hook(self, context):
         player_direction = self.pyboy.memory[
             self.pyboy.symbol_lookup("wSpritePlayerStateData1FacingDirection")[1]
@@ -1194,6 +1207,7 @@ def agent_stats(self, action):
                 "ptypes": self.read_party(),
                 "hp": self.read_hp_fraction(),
                 "coord": sum(sum(tileset.values()) for tileset in self.seen_coords.values()),
+                "warps": len(self.seen_warps),
                 "a_press": len(self.a_press),
                 "map_id": np.sum(self.seen_map_ids),
                 "npc": sum(self.seen_npcs.values()),
diff --git a/pokemonred_puffer/rewards/baseline.py b/pokemonred_puffer/rewards/baseline.py
index e16b864..5024828 100644
--- a/pokemonred_puffer/rewards/baseline.py
+++ b/pokemonred_puffer/rewards/baseline.py
@@ -330,6 +330,7 @@ def get_game_state_reward(self):
                 "saffron_guard": self.reward_config["required_event"]
                 * float(self.wd728.get_bit("GAVE_SAFFRON_GUARD_DRINK")),
                 "a_press": len(self.a_press) * self.reward_config["a_press"],
+                "warps": len(self.seen_warps) * self.reward_config["explore_warps"],
             }
             | {
                 f"exploration_{tileset.name.lower()}": self.reward_config.get(
diff --git a/pokemonred_puffer/wrappers/exploration.py b/pokemonred_puffer/wrappers/exploration.py
index 747a8ad..05a1427 100644
--- a/pokemonred_puffer/wrappers/exploration.py
+++ b/pokemonred_puffer/wrappers/exploration.py
@@ -60,6 +60,10 @@ def step_forget_explore(self):
             (k, max(0.15, v * (self.step_forgetting_factor["npc"])))
             for k, v in self.env.unwrapped.seen_npcs.items()
         )
+        self.env.unwrapped.seen_warps.update(
+            (k, max(0.15, v * (self.step_forgetting_factor["coords"])))
+            for k, v in self.env.unwrapped.seen_warps.items()
+        )
         self.env.unwrapped.explore_map *= self.step_forgetting_factor["explore"]
         self.env.unwrapped.explore_map[self.env.unwrapped.explore_map > 0] = np.clip(
             self.env.unwrapped.explore_map[self.env.unwrapped.explore_map > 0], 0.15, 1
@@ -113,6 +117,7 @@ def reset(self, *args, **kwargs):
             self.env.unwrapped.seen_npcs.clear()
             self.env.unwrapped.cut_coords.clear()
             self.env.unwrapped.cut_tiles.clear()
+            self.env.unwrapped.seen_warps.clear()
         self.counter += 1
         return self.env.reset(*args, **kwargs)
 
@@ -146,4 +151,9 @@ def reset(self, *args, **kwargs):
         self.env.unwrapped.cut_explore_map[self.env.unwrapped.cut_explore_map > 0] = (
             self.fixed_value["cut"]
         )
+        self.env.unwrapped.seen_warps.update(
+            (k, self.fixed_value["coords"])
+            for k, v in self.env.unwrapped.seen_warps.items()
+            if v > 0
+        )
         return self.env.reset(*args, **kwargs)