CORTEXA
← Browse
arxivcs.CLcs.LG2026-06-30

Explicit Fuzzy Logic in the Feed-Forward Layer: Self-Forgetting Quantifiers Discover Legible Grammatical-Licensing Detectors

Mark Oskin

A transformer's feed-forward (FFN) sublayer materializes the distinctions attention gathers, yet gives no account of what it computes. In a parameter-neutral replacement, each hidden unit is an explicit fuzzy set operation on sigmoid-bounded [0,1] memberships: intersection A*B and set-difference A*(1-B), the latter a bounded positive negation ("A but not B") that gated/bilinear units lack -- a negation-capable FFN (NC-FFN). On N-bit parity they are the most parameter-efficient reasoning basis at shallow depth; at scale (125M, OpenWebText) NC-FFN ties the GELU baseline's perplexity, every unit carrying explicit logical form. Two limits share one cause: two-operand logic localizes to layer 0 and erodes under training, and the one robust grammatical deficit concentrates in licensing and quantifiers, beyond within-token operators. We resolve both with a small block of sequence quantifiers: a soft existential and a soft proportion, each with a per-unit learned forgetting rate from a sticky init. This recovers the deficit at epoch one (halving the wider epoch-two gap), modestly leads on LAMBADA, and makes the FFN legible: the structure now holds and migrates into depth; the decay un-learns its stickiness (median half-life ~1.5 tokens; zero latch units); and at the semantic layers the units read, without dictionary learning, as grammatical licensing detectors: each fires on a licensor (a comparative, a passive participle, a negative-polarity item) and carries its memory forward to predict the licensed word (than, by, nor). This legibility is localized and free only up to a partition (a fully Boolean FFN diverges in training), but the result is a parameter-neutral, language-model-quality transformer with a readable, interpretable-by-construction grammatical mechanism -- an account not just of what a feed-forward layer represents but how it licenses.

View free PDFSource page

Related papers

arxivcs.LGcs.AIcs.CL2026-07-20

A Controlled Study of Attention-Only Transformers

Henry Ndubuaku, Karen Mosoyan, Jakub Mroz, Noah Cylich, Satyajit Kumar, Parkirat Sandhu, et al.

Feed-forward networks hold two thirds of a transformer's non-embedding parameters, yet the architecture has not received a necessity test that controls parameters, compute, and depth at once. We pretrain attention-only decoder transformers (Simple Attention Networks, SANs) agains…

View free PDFSource page
arxivcs.LGcs.CL2026-07-09

Training, Reading, and Editing Legible Transformers

Mark Oskin

A transformer can be built from operators that are legible by construction -- bounded, named units that read as fuzzy set operations rather than dense activations -- but legibility must be pressed for during training, and the pressure has a failure mode. A crispness penalty meant…

View free PDFSource page
arxivcs.CLcs.LG2026-07-08

Hallucination Self-Play: Bootstrapping Reinforced Detector via Evolved Generator

Shiping Yang, Shining Liang, Weihao Liu, Wenbiao Ding, Linjun Shou, Lu Cheng, et al.

Identifying faithfulness hallucinations in LLM-generated outputs remains challenging due to the scarcity of high-quality annotated data. Recent work relies on advanced LLMs to synthesize training data, including rationales, labels, and hallucinated claims. However, these methods…

View free PDFSource page
arxivcs.CLcs.AIcs.LG2026-07-03

CaresAI at SMM4H-HeaRD 2026: Predicting TNM Staging

Joseph Itopa Abubakar, Jorge Jarme, Favour Igwezeke, Mary Adewunmi

This study aims to predict Tumor, Node, and Metastasis (TNM) stage labels independently, with the Cancer Genome Atlas (TCGA) pathology report as the sixth shared task of SMM4H-HeaRD 2026. The problem is framed as three multi-label classification tasks. We explore both classical a…

View free PDFSource page
arxivcs.LGcs.AIcs.CLcs.CR2026-07-01

Black-Box Inference of LLM Architectural Properties with Restrictive API Access

Christopher Ellis, Shreyas Chaudhari, Mei-Yu Wang, Leighton Barnes, Giulia Fanti, José M. F. Moura

In practice, most commercial LLM providers do not publicly release details of underlying LLM architectures. However, prior work has shown that given limited API access to an LLM (namely, top-$k$ logits and/or a logit bias function), one can recover certain architectural details o…

View free PDFSource page