Split row cols by source: add rew_s/gt_t; rename timing col t_rew

The combined `rew` column mixed student + teacher rollouts, making it
hard to tell "is student learning?" at a glance. Add per-source splits:

- rew_s: student-only mean reward (primary learning signal)
- gt_t : teacher-only ground-truth pass count (cache stability check)

The teacher pool is frozen at startup (baseline logged at load), so
per-step rew_t adds noise without information and is omitted.

The previous `rew_s` column was actually reward-grading wall-time (an
unfortunate name collision with student reward). Rename it to `t_rew`
to match the other timing cols (gen, fb).

New column order:
  step ref_eq rew rew_s std sprd N
  gt gt_s gt_t hack hack_s hack_t
  loss cin cin_s cin_t cout fired
  gen fb t_rew sec

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

src/projected_grpo/train.py

@@ -485,7 +485,7 @@ def main(cfg: Config) -> int:
         f"SHOULD: loss finite each step; projected arm cos_out <= cos_in; "
         f"PASS_RATE > 0 on 4B (was 0/16 under broken grader). "
         f"ELSE: harness or projection broken. "
-        f"Timing cols (gen/fb/rew_s/sec): gen-bound -> vLLM; fb-bound -> lower pp; rew_s-bound -> parallel grading."
+        f"Timing cols (gen/fb/t_rew/sec): gen-bound -> vLLM; fb-bound -> lower pp; t_rew-bound -> parallel grading."
     )
     if teacher_pool:
         logger.info(
@@ -509,10 +509,15 @@ def main(cfg: Config) -> int:
     # So ref_eq=1.0 means we've issued the same number of gradient samples as
     # one canonical reference step. Convert our step count to "reference step
     # equivalents" by reading this column at the row of interest.
-    _row_cols = ["step", "ref_eq", "rew", "std", "sprd", "N",
-                 "gt", "hack", "hack_s", "hack_t", "gt_s",
+    # Per-source split (student/teacher) for rew, gt, hack columns. Teacher pool
+    # is frozen so rew_t/gt_t are mostly sanity checks that cache sampling is
+    # stable; rew_s/hack_s are the primary "is student learning?" signals.
+    # `t_rew` is the reward-grading wall-time (s); kept separate from `rew_s`
+    # (student mean reward) to avoid the name collision the older log had.
+    _row_cols = ["step", "ref_eq", "rew", "rew_s", "std", "sprd", "N",
+                 "gt", "gt_s", "gt_t", "hack", "hack_s", "hack_t",
                  "loss", "cin", "cin_s", "cin_t", "cout", "fired",
-                 "gen", "fb", "rew_s", "sec"]
+                 "gen", "fb", "t_rew", "sec"]
     REF_GENS_PER_STEP = 16 * 16  # ariahw/rl-rewardhacking config.py:num_prompts * num_generations
     est_gens_per_step = cfg.prompts_per_step * cfg.group  # before mixed-pool split
     logger.info(
@@ -840,6 +845,8 @@ def main(cfg: Config) -> int:
         hack_s_n = int((h_t & is_s).sum())
         hack_t_n = int((h_t & ~is_s).sum())
         gt_s_n   = int((g_t & is_s).sum())
+        gt_t_n   = int((g_t & ~is_s).sum())
+        rew_s_mean = rewards_t[is_s].mean().item() if n_s else float("nan")
 
         # Per-step diagnostics → verbose log; stdout sees tqdm postfix + final table.
         n_fin = sum(agg_finished)
@@ -869,14 +876,16 @@ def main(cfg: Config) -> int:
             "step": step,
             "ref_eq": f"{cum_gens / REF_GENS_PER_STEP:.2f}",
             "rew": f"{rew_mean:+.2f}",
+            "rew_s": f"{rew_s_mean:+.2f}" if n_s else "nan",
             "std": f"{rew_std:.2f}",
             "sprd": "T" if spread else "F",
             "N": n_rollouts,
             "gt": f"{sum(agg_gt)}/{n_rollouts}",
+            "gt_s": f"{gt_s_n}/{n_s}" if n_s else "0/0",
+            "gt_t": f"{gt_t_n}/{n_t}" if n_t else "0/0",
             "hack": f"{sum(agg_hack)}/{n_rollouts}",
             "hack_s": f"{hack_s_n}/{n_s}" if n_s else "0/0",
             "hack_t": f"{hack_t_n}/{n_t}" if n_t else "0/0",
-            "gt_s": f"{gt_s_n}/{n_s}" if n_s else "0/0",
             "loss": f"{agg_loss:+.4f}",
             "cin": f"{diag['mean_cos_in']:+.3f}",
             "cin_s": f"{diag['mean_cin_s']:+.3f}",
@@ -885,7 +894,7 @@ def main(cfg: Config) -> int:
             "fired": f"{diag['frac_fired']:.2f}",
             "gen": f"{t_gen:.0f}",
             "fb": f"{t_fb:.0f}",
-            "rew_s": f"{t_rew:.0f}",
+            "t_rew": f"{t_rew:.0f}",
             "sec": f"{time.time()-t0:.0f}",
         }
         rows.append(row)