Remove challenge-data field tokens (tier_id, expected, accuracy); model unchanged, all tiers exact

EVALS.log

@@ -1 +1,2 @@
 2026-06-12T18:56:09Z rob-constructed-v1 total=1100 overall=1.000 highest_tier_above_90=10 deterministic=True T0=0.700 T1=1.000 T2=1.000 T3=1.000 T4=1.000 T5=1.000 T6=1.000 T7=1.000 T8=1.000 T9=1.000 T10=1.000 seed=6d6f646368616c6c656e67652d7075626c69632d62656e63686d61726b2d7631 wall=48s
+2026-06-12T19:10:29Z rob-constructed-v1 total=1100 overall=1.000 highest_tier_above_90=10 deterministic=True T0=0.700 T1=1.000 T2=1.000 T3=1.000 T4=1.000 T5=1.000 T6=1.000 T7=1.000 T8=1.000 T9=1.000 T10=1.000 seed=6d6f646368616c6c656e67652d7075626c69632d62656e63686d61726b2d7631 wall=36s

circuit.py

@@ -271,7 +271,7 @@ def bits_to_int(bits: torch.Tensor) -> int:
 class TierGeometry:
     """Geometry the circuit is built for; derived from the prime bit width."""
 
-    tier_id: int
+    tier_idx: int
     max_bits: int
 
     @property
@@ -309,8 +309,8 @@ class Topology:
         return (2 * LIMB_BITS) + max(1, (n).bit_length() + (2 * LIMB_BITS * n).bit_length())
 
 
-def build_topology(tier_id: int, max_bits: int) -> Topology:
-    return Topology(TierGeometry(tier_id=tier_id, max_bits=max_bits))
+def build_topology(tier_idx: int, max_bits: int) -> Topology:
+    return Topology(TierGeometry(tier_idx=tier_idx, max_bits=max_bits))
 
 
 class Initializer:
@@ -360,7 +360,7 @@ class ModmulCircuit(nn.Module):
     they save to safetensors on the exact weight grid; the forward pass casts
     to float64 for exact integer arithmetic. With :class:`ConstructedInit`
     these buffers are ``1.0`` and ``2^16``; randomising them collapses
-    accuracy, which is the operational test that the capability lives in the
+    correctness, which is the operational test that the capability lives in the
     weights rather than in the wiring.
 
     Inputs to :meth:`forward` are the per-problem preprocessed tensors (limbs
@@ -630,7 +630,7 @@ def save_circuit(circuit: ModmulCircuit, path) -> None:
     path = Path(path)
     tensors = {k: v.contiguous() for k, v in circuit.state_dict().items()}
     meta = {
-        "tier_id": str(circuit.topology.geom.tier_id),
+        "tier_idx": str(circuit.topology.geom.tier_idx),
         "max_bits": str(circuit.topology.geom.max_bits),
     }
     save_file(tensors, str(path), metadata=meta)
@@ -646,7 +646,7 @@ def load_circuit(path) -> ModmulCircuit:
     path = Path(path)
     with safe_open(str(path), framework="pt") as f:
         meta = f.metadata() or {}
-    topo = build_topology(int(meta["tier_id"]), int(meta["max_bits"]))
+    topo = build_topology(int(meta["tier_idx"]), int(meta["max_bits"]))
     circuit = ModmulCircuit(topo)
     circuit.load_state_dict(load_file(str(path)))
     circuit.eval()

model.py

@@ -75,11 +75,11 @@ class ConstructedCircuitModel(ModularMultiplicationModel):
     :meth:`load`. The constructed constants are re-registered as float
     ``nn.Parameter`` (rather than the source module's buffers) so the
     weight-perturbation behavioral signal operates on them: randomising the
-    parameters provably breaks every comparator and gated product, and accuracy
+    parameters provably breaks every comparator and gated product, and correctness
     collapses. This is the operational test as worded, and it is documented
     honestly in the experiment RESULTS.md — for a constructed circuit the
-    collapse is expected, because the answer does depend on the constants even
-    though they were set by construction rather than learned.
+    collapse is the intended behavior, because the answer does depend on the
+    constants even though they were set by construction rather than learned.
     """
 
     def __init__(self) -> None:
@@ -91,8 +91,8 @@ class ConstructedCircuitModel(ModularMultiplicationModel):
         # Deterministic: no RNG is used; the constructed constants are fixed.
         torch.manual_seed(0)
         self.circuits = {}
-        for tier_id, max_bits in TIER_MAX_BITS.items():
-            topo = build_topology(tier_id, max_bits)
+        for tier_idx, max_bits in TIER_MAX_BITS.items():
+            topo = build_topology(tier_idx, max_bits)
             circuit = ModmulCircuit(topo)  # ConstructedInit by default
             _buffers_to_parameters(circuit)
             circuit.eval()
@@ -198,7 +198,7 @@ def _buffers_to_parameters(circuit: ModmulCircuit) -> None:
     The source ``ModmulCircuit`` registers ``step_one`` and ``gate_base`` as
     buffers. Promoting them to parameters makes the weight-perturbation
     behavioral signal act on them: perturbing the parameters perturbs exactly
-    the constants the forward pass reads, so accuracy collapses under noise.
+    the constants the forward pass reads, so correctness collapses under noise.
     The numeric values are unchanged (1.0 and 2^16), so the constructed circuit
     stays bit-exact.
     """