remove full pos embedding

src/lerobot/policies/rlearn/configuration_rlearn.py

@@ -80,7 +80,7 @@ class RLearNConfig(PreTrainedConfig):
     # ReWiND augmentation
     rewind_prob: float = 0.3 #0.8
     rewind_last3_prob: float = 0.0 #0.3
-    mismatch_prob: float = 0.0# 0.2
+    mismatch_prob: float = 0.0 #0.2
     
     # Normalization presets
     normalization_mapping: dict[str, NormalizationMode] = field(

src/lerobot/policies/rlearn/modeling_rlearn.py

@@ -86,10 +86,6 @@ class RLearNPolicy(PreTrainedPolicy):
 
         # First-frame positional embedding (only applied to the first video frame)
         self.first_frame_pos = nn.Parameter(torch.zeros(1, 1, config.dim_model))
-        # Full temporal positional embeddings (length = max_seq_len)
-        self.max_time = config.max_seq_len
-        self.time_pos = nn.Parameter(torch.zeros(1, self.max_time, config.dim_model))
-        nn.init.trunc_normal_(self.time_pos, std=0.02)
 
         # Cross-modal sequential aggregator – causal transformer over
         # [language tokens | video frame tokens] using PyTorch TransformerEncoder
@@ -330,9 +326,7 @@ class RLearNPolicy(PreTrainedPolicy):
         # SigLIP2 CLS per-frame already returned
         video_frame_embeds = video_patch_embeds  # (B, T_eff, D_vision)
         video_tokens = self.to_video_tokens(video_frame_embeds)  # (B, T_eff, D)
-        # Add temporal positional embeddings
-        video_tokens = video_tokens + self.time_pos[:, :T_eff, :]
-        # Optional: keep a first-frame tag
+        # ReWiND: only add a first-frame positional bias (prevents time cheating)
         video_tokens[:, :1, :] = video_tokens[:, :1, :] + self.first_frame_pos
 
         # Build masks for TransformerEncoder