A coding agent can fail in two very different ways. One failure is obvious: it does not think enough. It sees an error report, guesses the wrong file, edits too early, and then spends the rest of the trajectory debugging its own mistake. Anyone who has watched an autonomous coding agent wander through a repository has seen this little tragedy. The machine is busy, but not necessarily useful. ...
Spreadsheet errors have a special talent: they look boring until they become expensive. That is the business version of the LLM math problem. A model can produce a calm, step-by-step explanation, put a confident number at the bottom, and still be wrong in the only place that matters. Worse, the reasoning may look plausible enough that a manager, analyst, tutor, or compliance reviewer nods and moves on. The answer has the rhythm of thinking. It has the costume of calculation. It may even have a chain-of-thought trace. Very civilized. Still not proof. ...
Think Longer, Act Worse? What M2A Teaches About Reasoning Agents A coding agent does not fail only because it cannot think. Sometimes it fails because it keeps thinking after it should inspect the repository. Sometimes it writes a plausible explanation before checking the relevant file. Sometimes it burns the context window by wandering through hypotheses, each one almost reasonable, none of them decisive. The result is not stupidity in the familiar sense. It is a coordination failure: the model does not know when to reason, when to call a tool, when to absorb feedback, and when to edit. ...
Opening — Why this matters now The AI industry has discovered a wonderfully pedestrian way to misread progress: build models that can solve harder math problems, then grade them with evaluators that panic when 2040 minutes is not written as 34 hours. That is not a joke. It is the central irritation behind “Rethinking Math Reasoning Evaluation: A Robust LLM-as-a-Judge Framework Beyond Symbolic Rigidity”, an arXiv paper that examines how mathematical reasoning benchmarks can be distorted by rigid symbolic verification.1 ...
Search. That is the unglamorous part of AI work. The demo asks a model to solve a clean problem. The enterprise system asks a model to find the right prior case, retrieve the relevant precedent, avoid the misleading near-match, and then adapt the answer without making a confident mess of it. MathNet is interesting because it puts that distinction under pressure. The paper introduces a large multilingual, multimodal Olympiad mathematics benchmark, but the more useful business lesson is not merely that frontier models can solve hard math. We already have enough leaderboards wearing medals. The sharper finding is that models and embedding systems can still fail at recognizing when two problems are mathematically the same, or when one problem is structurally useful for another.1 ...
Most AI math benchmarks still feel like exam rooms. The model receives a problem. It produces an answer. We score the answer. Everyone argues about whether the problem was hard enough, whether the model saw something similar during training, and whether the leaderboard means anything outside the leaderboard. Very productive. Almost as peaceful as a faculty meeting. ...
Proof is where AI systems stop performing confidence and start owing the reader money. A model can restate a theorem elegantly. It can cite the right neighborhood of literature. It can produce LaTeX with the visual manners of a publishable paper. None of that is a proof. It is proof-shaped material. Sometimes useful. Sometimes impressive. Sometimes a very expensive fog machine. ...
A coder does not trust a program because it sounds plausible. A coder runs it, reads the error message, changes the implementation, tests again, searches the library, asks a colleague, splits the problem, and keeps going until the machine stops complaining. That mundane loop is the interesting part of Numina-Lean-Agent: An Open and General Agentic Reasoning System for Formal Mathematics.1 The headline result is easy to market: with Claude Opus 4.5 as the base model, Numina-Lean-Agent solves all 12 Putnam 2025 problems in Lean, matching the reported perfect score of AxiomProver. Nice. The trophy cabinet sparkles. ...
The invoice arrives after the benchmark party Math benchmarks are fun until the training bill arrives. A model can be taught to produce longer reasoning traces. It can be shown more olympiad problems. It can be given Python. It can be pushed into 128K-token contexts and told, heroically, to think harder. All of this sounds impressive in a benchmark table. Less impressive is the operational detail that most training samples do not need the full 128K window, yet a naive training setup can still make every step pay for it. ...
Review is not a democracy. That sounds unpleasant, which is why it is useful. In many business settings, we like consensus because it feels stable. Three analysts agree, five reviewers approve, the dashboard turns green, and everyone can pretend the risk has been domesticated. Mathematics is less polite. One invalid theorem application, one hidden assumption, one algebraic step that does not follow, and the whole proof may collapse. The majority does not get to vote a contradiction out of existence. ...