card: the formal-logic A/B against the base - 10/12 vs 9/12, no difference detected

README.md

@@ -29,7 +29,17 @@ library_name: litert-lm
 
 Graded with `--max-tokens 2048`: this is a reasoning model and emits a `<think>` block before answering, so a smaller budget scores it as wrong when it has simply not finished thinking.
 
-**Not measured: the formal-logic claim.** Upstream describes TwIL-LM3 as specialised for formal logic. The eight-question gate is general knowledge, so it cannot see that claim in either direction — a model could be better or worse at logic than its base and score identically here. This card therefore makes **no claim about logical reasoning ability**, and none should be read into the 8/8.
+### The formal-logic claim: measured against the base, and no difference detected
+
+Upstream describes TwIL-LM3 as specialised for formal logic. The eight-question gate is general knowledge and cannot see that claim in either direction, so it was tested separately: **12 items with one defensible answer each** (syllogisms, modus tollens, affirming the consequent, contrapositive, De Morgan, quantifier negation, disjunctive syllogism, transitivity, necessary vs sufficient, biconditional, vacuous truth), run against **this model and its base [SmolLM3-3B](https://huggingface.co/HuggingFaceTB/SmolLM3-3B) on identical items, identical flags, identical runtime, and identically-converted bundles** (same rail, same recipe, both 2.00 GB, both duplication-free).
+
+| | TwIL-LM3 | SmolLM3-3B (base) |
+|---|---|---|
+| score | 10 / 12 | 9 / 12 |
+
+**Read this as "no difference detected", not as a win.** The two models disagree on only **3 of the 12 items**, splitting 2–1 — TwIL-LM3 is right on the valid syllogism and the contrapositive where the base is wrong, the base is right on De Morgan where TwIL-LM3 is wrong, and both fail the vacuous-truth item. A 2–1 split among three discordant items is indistinguishable from chance, and a 12-item instrument can only detect large effects in any case. The honest summary is that **this test was not sensitive enough to confirm or refute the specialisation**, and the one-point total should not be quoted as evidence for it.
+
+The harness is `scripts/eval_logic.py` in the [converter repo](https://github.com/john-rocky/hf-to-litertlm), so the comparison can be repeated or extended.
 
 ## Performance