diff --git a/src/game/engine/src/modules/management/turn_processor.gd b/src/game/engine/src/modules/management/turn_processor.gd
index bd581238..df6ea65f 100644
--- a/src/game/engine/src/modules/management/turn_processor.gd
+++ b/src/game/engine/src/modules/management/turn_processor.gd
@@ -153,7 +153,15 @@ func _process_research(player: RefCounted) -> void:  # Player
 		return
 
 	# Per-yield difficulty multiplier (composed by GameState).
+	# p1-29 cycle 4: research-yield catch-up boost. When this player is at least
+	# 2 eras behind the highest opponent on the tech ladder, multiply science
+	# yield by 1.5× — composes multiplicatively with cycle-3's tech-PICK boost in
+	# auto_play.gd::_pick_research. Cycle 3 batch confirmed pick-side multiplier
+	# alone doesn't move tier_peak_gap because the loser is research-OUTPUT
+	# starved, not just choosing wrong techs. The output boost addresses that
+	# directly. Inert when both players are at similar eras (`is_behind=false`).
 	var sci_modifier: float = GameState.get_effective_yield_mult(player, "research")
+	sci_modifier *= _catchup_research_mult(player)
 
 	# Per-city science yields the Rust side will sum.
 	var game_map: RefCounted = GameState.get_game_map()
@@ -497,3 +505,64 @@ static func _player_owns_resource(player: RefCounted, resource_id: String) -> bo
 			if tile != null and str(tile.get("resource_id")) == resource_id:
 				return true
 	return false
+
+
+# ── p1-29 cycle 4: research-output catch-up dynamics ─────────────
+#
+# Helpers below mirror the static functions added to auto_play.gd in cycle 3
+# (`_player_tier_peak`, `_max_opponent_tier_peak`). Duplicated here rather than
+# referenced via AutoPlayScript so that turn_processor's _process_research path
+# is self-contained — auto_play.gd is an autoload that may not be loaded in
+# every code path that ends up calling this module (live games, GUT tests,
+# headless MCTS).
+
+
+static func _player_tier_peak(player: RefCounted) -> int:
+	## Compute the highest tech-era this player has researched. Mirrors the
+	## per-player tier_peak metric written into turn_stats.jsonl.
+	if player == null:
+		return 0
+	var techs: Array = player.researched_techs if player.get("researched_techs") != null else []
+	var peak: int = 0
+	for tid: Variant in techs:
+		var data: Dictionary = DataLoader.get_tech(str(tid))
+		var era: int = int(data.get("era", 0))
+		if era > peak:
+			peak = era
+	return peak
+
+
+static func _max_opponent_tier_peak(self_player: RefCounted) -> int:
+	## Compute the highest opponent's tier_peak across all OTHER players in
+	## GameState.players. Returns 0 when self is the only player or no
+	## opponents have researched anything yet.
+	var max_peak: int = 0
+	if GameState == null or GameState.players == null:
+		return 0
+	var self_idx: int = -1
+	if self_player != null and self_player.get("index") != null:
+		self_idx = int(self_player.index)
+	for p: Variant in GameState.players:
+		if p == null:
+			continue
+		var pi: int = int(p.index) if p.get("index") != null else -1
+		if pi == self_idx:
+			continue
+		var peak: int = _player_tier_peak(p)
+		if peak > max_peak:
+			max_peak = peak
+	return max_peak
+
+
+static func _catchup_research_mult(player: RefCounted) -> float:
+	## Returns 1.5× when this player is at least 2 eras behind the highest
+	## opponent's tech ceiling, 1.0× otherwise. Composes multiplicatively with
+	## the configured base research multiplier (set by AI_DIFFICULTY etc.) and
+	## with auto_play.gd::_pick_research's pillar-multiplier boost. (p1-29 cycle 4)
+	if player == null:
+		return 1.0
+	var self_tp: int = _player_tier_peak(player)
+	var opp_tp: int = _max_opponent_tier_peak(player)
+	if opp_tp >= self_tp + 2:
+		return 1.5
+	return 1.0