arXiv subjects

Guanghui Wang

Publications and source records attributed to Guanghui Wang.

At least 19 recordsLinked to original sources

arXiv2026

Limiting Pointwise Decay for the compressible isentropic Navier-Stokes equations

We study the long-time pointwise behavior of small localized perturbations of a constant state for the one-dimensional compressible isentropic Navier-Stokes equations. After subtracting the two Burgers diffusion waves, the convergent sum of all higher-order diffusion waves, and the cross-family viscous corrections, we prove a cone-preserving pointwise estimate for the exact physical remainder and, in particular, \[ |R_i(x,t)|\leq C E_N\log(2+t)Ψ_i(x,t), \qquad \sup_{x\in\mathbb R}Ψ_i(x,t)\leq C(1+t)^{-1}. \] Here $E_N$ measures the size of the initial data and $Ψ_i$ is the cone-resolved weight; both are defined precisely in the main theorem below. Thus $\|R_i(t)\|_{L^\infty}\leq C E_N(1+t)^{-1}\log(2+t)$. The key new idea is to apply a familywise Cole--Hopf transformation to the spatial antiderivative of the remainder, which exactly eliminates the critical same-family first-order feedback. We further construct an approximate Green function adapted to the two characteristic families and combine it with Gaussian-mode extraction and a Kawashima-type energy argument. This yields a cone-preserving estimate at the limiting decay rate, up to a logarithmic loss.

math.AP↗

arXiv2026

UnitBoost: Managing Compound LLM Systems with a Merge Operator, Not a Model

Compound LLM systems often solve a coordination problem by adding a higher-level LLM. The resulting meta-agent reads workers' outputs, writes the final answer, allocates later calls, and decides when to stop. It is expressive, but it also concentrates three control decisions in an opaque, order-sensitive model call. We ask whether the manager needs to be generative at all. UnitBoost replaces that model with a defined meta-level operator: a task-given unit map turns worker outputs into slot-value proposals, a constrained argmax assembles the output, and the slots left unfilled or unsupported become an explicit residual for the next round. The operator is order-free, records unit provenance, and gives a simple guarantee: without coupling constraints, unit-wise maximization under the same admission score dominates selection of any complete candidate. On three held-out benchmarks, it exceeds the best single candidate chosen with gold labels by 0.060-0.195 absolute task-score points and input-matched generative managers by 0.048-0.076. Replacing only the management step improves six compound-system configurations by 0.013-0.182. Residual-directed rounds raise FanOutQA cell F1 from 0.4778 to 0.5524; matched controls show that the true residual outperforms random targets and ordinary rereading, while a label-free supply signal flags exhaustion after one unproductive round. The same analysis measures three conditions in which no such gain is available (one indivisible unit, unavailable unit identity, and an endpoint that charges for every emitted unit) and quantifies cross-unit coupling as a repair cost. The manager gives up semantic freedom and gains order invariance, unit provenance, and testable failure conditions.

cs.AI↗

arXiv2026

A Finite-Entropy Criterion for the Entropic Conditional Central Limit Theorem

We prove a finite-entropy criterion for the entropic conditional central limit theorem. Let $(ξ_i,η_i)_{i\geq 1}$ be independent copies of a pair $(ξ,η)$, and set $W_n=n^{-1/2}\sum_{i=1}^n ξ_i$ and $\boldsymbolη_n=(η_1,\ldots,η_n)$. Under the assumptions that $\mathbb{E}\operatorname{Var}(ξ\midη)<\infty$ and that the conditional law of $ξ$ given $η$ is absolutely continuous almost surely, we show that $\mathbb{E}h(W_n\mid\boldsymbolη_n)$ converges to the Gaussian entropy $\frac12\log(2πeσ^2)$, where $σ^2=\mathbb{E}\operatorname{Var}(ξ\midη)$, if and only if $\mathbb{E}h(W_{n_0}\mid\boldsymbolη_{n_0})>-\infty$ for some $n_0$. The main technical ingredient is a continuity theorem for Fisher information under Gaussian smoothing, which allows us to replace the finite expected conditional Fisher-information assumption by a necessary and sufficient finite-entropy condition.

math.PR↗

arXiv2026

Beyond Isolated Heads: Multi-Overlapped-Head Self-Attention for Vision Transformers

Multi-Head Self-Attention (MHSA) is the cornerstone of Vision Transformers, allowing models to capture diverse feature representations by projecting tokens into independent subspaces. However, the standard MHSA strictly isolates these heads, preventing any information exchange during the attention computation itself. In this paper, we propose Multi-Overlapped-Head Self-Attention (MOHSA), a novel mechanism that replaces the hard division of attention heads with a soft, overlapping division. By allowing queries, keys, and values to partially overlap with those of adjacent heads, MOHSA fosters rich inter-head communication directly within the attention mechanism. We extensively investigate various overlap dimension scheduling strategies to identify optimal configurations. Comprehensive experiments in several Transformer architectures demonstrate that MOHSA outperforms standard MHSA on the CIFAR-10, CIFAR-100, Tiny-ImageNet, and ImageNet-1k datasets, offering a significant performance boost with negligible computational overhead. Code link: https://github.com/ZTX-100/MOHSA.

cs.CV↗

arXiv2026

Metrics That Write Themselves: Evolving an Evaluator from Its Own Blind Spots

Agents improve quickly against a reliable automatic metric and stall without one, and the applications that need them most, report generation among them, are the ones nobody knows how to score. Can the metric write itself? Saying what makes an answer good is hard; pointing at something wrong with one is easier, so the metric we evolve is a pool of small Python operators that each flag a candidate for one named defect, or abstain, and vote. Asking a model for operators directly does not work: 183 candidates realise only 96 distinct behaviours, from one narrow region of an enormous space. EvalCEGAR instead borrows counterexample-guided abstraction refinement from program verification. It reads the pool as an abstraction and searches for a collision, two answers the operators score identically, one correct and one not. That pair, not a prompt, is the authoring request, and when a collision defeats every attempt the loop widens what an operator may read rather than resampling. On MBPP+ and HumanEval+, a sandbox whose hidden unit tests give exact ground truth, the loop writes a 55-line operator that closes 15.4% of the gap between flagging nothing and a perfect filter on 428 unseen tasks (+0.0065, p=0.0010) at a quarter of our best hand-written operator's flags. On the benchmark it never saw it matches that operator's effect exactly on a third of the flags. Six of eight runs admit such an operator and all six help out of sample; our 15 hand-written operators applied together as one filter lose accuracy. An LLM judge on the same information ties that delta on a nearly disjoint set of candidates, and charges a model call per candidate forever where the operator charges none.

cs.AI↗

arXiv2026

The Blind Curator: How a Biased Judge Silently Disables Skill Retirement in Self-Evolving Agents

A self-evolving agent retires its bad skills by watching them fail, so what happens when the judge cannot see the failures? Skill retirement is the structural constraint that keeps a growing library from drifting below the no-skill baseline, but its guarantee assumes an unbiased reward, which is false for the LLM judges that reference-free tasks require. We show that a biased judge does not merely add noise; it \emph{silently switches off the curator}. We make this precise with a corrupted-reward analysis, then a behavioral study on a reference-free report-writing testbed with a code-generation cross-check, injecting corruption on top of a deterministic reward to isolate the causal channel. Symmetric noise leaves retirement intact, but \emph{false-pass} bias (failures slipping through as passes) disables contribution-based retirement past a sharp threshold (here a false-pass rate of $0.45$) that no amount of data can cross. Separating genuine retirement from cap-eviction churn shows this \emph{mechanism} failure is universal, holding across domains and failure rates and sparing only near-zero-false-pass, verifier-like graders. The downstream \emph{outcome}, though, is regime-dependent: eval quality degrades only where the same corruption also starves skill synthesis, and otherwise holds steady, so the disabled curator is \emph{silent}, surfacing in no aggregate metric. The contribution is a behavioral safety result, not a performance one. A cheap defect-injection audit then tells an operator, before deployment, which side of the threshold their judge occupies.

cs.AI↗

arXiv2026

Extracting a nitrile-centered, ether-assisted motif hierarchy for lithium-battery electrolyte design from billion-scale molecular space

Designing electrolyte molecules for lithium batteries requires balancing electronic stability with appropriate Li+ solvation, yet the structural basis remains unclear across chemically diverse molecules. High-throughput screening expands the searchable space, but ranked candidates alone do not reveal recurring motifs or their applicability limits. We searched nearly one billion GDB13 structures using electronic--solvation descriptors without explicit functional-group preferences or scaffold constraints. Across descriptor weights, high-ranking populations separated into a nitrile-dominant regime and a coexistence regime containing substantial fractions of both nitrile- and ether-containing molecules. These regimes together define a nitrile-centered, ether-assisted motif hierarchy: nitrile remains favored across broad weight ranges, whereas ether becomes prominent under stronger electrostatic and polarity constraints. Encoding this hierarchy in a generative model expands the candidate space beyond GDB13 and yields high-scoring fluorinated structures without an explicit fluorination reward. Explicit-solvent molecular dynamics simulations show weak, exchangeable coordination of representative candidates without displacing ethylene carbonate from the dominant first solvation shell around Li+; effects on ion association and transport depend on molecular structure and concentration. These results establish a quantitative, interpretable and physically bounded motif hierarchy that systematizes established nitrile and ether chemistry for lithium-battery electrolyte design.

physics.chem-ph↗

arXiv2026

Reconcile Once, Write Anytime: A Trust-Tiered Librarian and a Multi-Agent Writer for Drift-Free, Point-in-Time Research

Long-form research reports generated by large language models drift, contradict themselves, and lose provenance: the same metric appears with different values, and rumor is quoted as confidently as an audited filing. We present a two-tier agentic system that separates a maintained, point-in-time knowledge library from report writing. A deterministic "librarian" ingests timestamped sources into a trust-tiered ontology, layering evidence cards, an authoritative metric ledger, and a claim graph into an always-current source of truth, not per-query RAG over raw chunks. A portable multi-agent "writer" runtime then composes a contradiction-free, evidence-grounded report at any knowledge cutoff T, reading only evidence with as_of <= T (no look-ahead); red-team verdicts flow back into the librarian. We evaluate on a self-collected, public corpus of 6,130 sources yielding 555,926 evidence cards (SEC EDGAR filings across 295 issuers and 11 sectors, U.S. Bureau of Labor Statistics releases, and Wikipedia). From the one library we compose four point-in-time reports on distinct theses and run eight reproducible experiments, whose headline metrics come from a deterministic quality-control gate, itself validated by defect-injection meta-evaluation at recall 1.0 and precision 1.0. A shared metric ledger removes 6,845 cross-section contradictions to zero. Tier-first selection is correct on 22/22 gold cases where a popularity-first baseline scores only 9/22; trust tiering leaks zero media-sourced numbers, and no government statistic displaces a company's own filing. A red-team refutation propagates back and self-corrects a later run with zero manual edits. Replay exhibits zero look-ahead violations across seven cutoffs while the library grows from 235,373 to 555,312 cards. Difficulty-tiered model routing exceeds the all-Opus quality ceiling while running 3.7x faster than serial.

cs.MA↗

arXiv2026

Ratchet: How Reliable Must an LLM Judge Be to Retire a Skill?

A large language model (LLM) agent that writes and edits its own skill library must also decide which skills to keep, from one noisy scalar per skill. The answer is exact: a judge scoring failures as passes at rate $(1-τ)/2$ or above retires nothing, at any sample size, for eviction margin $τ$. Audits find that machinery is rarely built: LLM-written skills are worth $+0.0$ percentage points (pp) against a no-skill control, human-written ones $+16.2$pp. Unmaintained, a library enters \emph{library drift}, growing until injecting a skill scores worse than injecting nothing. \textbf{Ratchet} repairs this: it evicts each skill on its measured contribution, caps the library at width $C$, and constrains synthesis, lifting held-out $pass@1$ by $+0.328$ on a hard MBPP+ slice. The matching non-divergence bound is finite for exactly two reasons, $C$ and $τ$. Our contribution is the condition this repair carries and no deployed system states. In reference-free domains the scalar comes from an LLM judge, whose two error directions, modelled as a binary channel, behave nothing alike. Passes scored as failures cost sample efficiency, which more trials buy back; failures scored as passes displace the eviction statistic, and no correction inside the rule recovers it. End-task score is a poor alarm, moving by at most a fifth of the governed lift and not monotonically in the rate. We prove both edges of the certifiable region, confirm them in a running loop, and place a judge on a known side in one offline pass.

cs.AI↗

arXiv2026

Who Grades the Grader? Co-Evolving Evaluation Metrics and Skills for Self-Improving LLM Agents

Self-evolving agent systems create, revise, and retire their own skills, but every such loop assumes a reliable evaluation metric already exists. In many real applications none does. We show the metric itself can be the evolving object: our loop searches compositions of small typed drawback detectors under a full evolutionary lifecycle, selecting for agreement with a ten-item anchored reference set and regularizing by consensus over unlabeled outputs. What evolves is the function that grades one output, never the fixed task sets it is scored on, and what comes out is an inspectable expression rather than an opaque judge. It is also valid: on code generation it gains 0.21 agreement with hidden ground truth on a locked set that metric selection never reads (paired $p=0.014$), beating the bare LLM judge it contains. Validity is where safety lives: removing the anchor guards collapses the metric into a vacuous always-pass detector while removing the detector lifecycle does not, inverting the lesson from skill evolution. That collapse warns this line of work that downstream task score cannot validate a self-evolved evaluator, since the collapsed metric trains skills just as well. Task score answers only sufficiency, and an evolved metric suffices: \emph{Double Ratchet}, co-evolving the metric with a lifecycle-managed skill loop, retains 88--110\% of the lift ground truth or a hand-written rubric buys, across MBPP+, Spider~2.0-Snow, and report generation. When evolved skills gamed the report rubric, an independent judge caught it and one added detector repaired it.

cs.AI↗

arXiv2026

Library Drift: Diagnosing and Fixing a Silent Failure Mode in Self-Evolving LLM Skill Libraries

Self-evolving skill libraries face a silent failure mode we term \emph{library drift}: unbounded skill accumulation without outcome-driven lifecycle management causes retrieval degradation, false-positive injections, and performance stagnation. Recent evaluation confirms the symptom (LLM-authored skills deliver +0.0pp gain while human-curated ones deliver +16.2pp (SkillsBench)), yet the underlying mechanism has not been isolated. We provide (1) a reproducible trigger: ablations that isolate drift: one disables skill injection (flat floor, +0.002), one imposes premature retirement (active harm, $-$0.019); (2) trace-level diagnostics: an append-only evidence log with per-skill contribution scores, attribution verdicts, and router engagement metrics that make the failure visible before it reaches end-task scores; and (3) a verified fix: a minimal governance recipe (outcome-driven retirement + bounded active-cap + meta-skill authoring prior) that lifts held-out pass@1 from a 0.258 baseline to a late-window mean of 0.584 (rolling gain $+$0.328) on MBPP+ hard-100 over 100 rounds. Eight ablations decompose which governance mechanisms are load-bearing and which are subsumed, providing a concrete playbook for diagnosing library drift in any self-evolving agent.

cs.AI↗

arXiv2026

MATS: A novel multi-modality multi-task learning framework for 3D perception in autonomous driving

Multi-modality data from different sensors provides rich complementary information for 3D perception, becoming an essential component in reliable autonomous driving systems. Current research typically designs intricate and complex fusion strategies to integrate information from multimodal data on a unified bird's-eye-view (BEV) feature map for the joint learning of multiple perception tasks. However, such a single feature map hardly carries sufficient information to simultaneously meet the requirements of various perception tasks, leading to a very limited perception performance. To mitigate this limitation, this paper proposes MATS, a novel multi-modality multi-task learning approach with modality-adaptive BEV fusion and task-specific Mixture-of-Experts (MoE) for 3D perception. Specifically, a simple modality-adaptive BEV fusion module is designed to adaptively recalibrate the BEV features by modeling the global cross-modality dependencies, generating diverse BEV feature maps for various perception tasks. For joint multi-task learning, this paper proposes a task-specific MoE module to decouple the tasks and enable the network to automatically choose the appropriate BEV feature candidates for each specific task. To validate the effectiveness of the proposed approach, we conduct extensive experiments on the large-scale benchmark nuScenes. With the camera- and LiDAR-modality input data, the proposed approach outperforms the state-of-the-art (SOTA) by a significant margin. Furthermore, the experimental results on the single tasks show that the proposed approach significantly outperforms the baselines. The code and trained models will be available upon publication.

cs.CV↗

arXiv2026

Sharp Diagonal Thresholds for Tight Hamilton Cycles in Uniformly Dense $3$-Graphs

A $3$-uniform hypergraph (or $3$-graph) $H$ on $n$ vertices is \emph{$(n,d,μ)$-dense} if $e_H(X,Y,Z)\ge d|X||Y||Z|-μn^3$ for all $X,Y,Z\subseteq V(H)$. This is one of the weakest standard notions of quasirandomness for $3$-graphs and is also known as linear quasirandomness. In this paper, we determine the sharp diagonal thresholds for tight Hamilton cycles in $(n,d,μ)$-dense $3$-graphs $H$ under conditions on the minimum vertex degree $δ_1(H)$ and the minimum codegree $δ_2(H)$. We actually prove a general result: define \[ f(d):=\frac{1-\sqrt{(4d-1)/3}}2. \] We prove that $(n,d,μ)$-density together with $δ_1(H)\geα\binom{n-1}{2}$ forces a tight Hamilton cycle whenever $d > 1/3$ and $α>f(d)$. In particular, $f(1/3)=1/3$, which answers Problem~8.3(i) of Araújo, Piga and Schacht and confirms Conjecture~8.1 of Han, Shu and Wang. For the minimum codegree condition, the sharp diagonal threshold is $(κ,κ)$, where $κ$ is the unique real solution of $κ=(1-κ)^3$. Since $κ\approx0.3177>1/4$, this gives a negative answer to Problem~8.3(ii) of Araújo, Piga and Schacht and disproves Conjecture~8.2 of Han, Shu and Wang. The two proofs use a common Hamilton-framework reduction, but the two degree conditions lead to distinct dominant-component lemmas for $(n, d, μ)$-dense $3$-graphs, which are of independent interest and whose proofs do not rely on the absorption method.

math.CO↗

arXiv2026

Dynamic Layered Decoding Scheduling for LDPC Codes Aided by Check Node Unsatisfied Probabilities

This letter revisits update ordering in layered belief propagation (LBP) decoding of low-density parity-check (LDPC) codes. The closest probability-based schedule orders layers by check node unsatisfied probabilities estimated only from the channel observations, although these probabilities change once decoding messages are exchanged. We therefore refresh the check node unsatisfied probabilities during decoding and use them as dynamic priorities. The first schedule, Dyn-EBP, selects the most reliable available check node while ensuring that every check node is updated once in each iteration. The second schedule, Dyn-PEBP, adds a linear update-count penalty and allows limited repeated updates without letting a small subset of check nodes dominate the schedule. For 5G new radio LDPC base graph 1 codes with five iterations, the proposed schedules yield small BLER reductions relative to the channel-only probability schedule and remain competitive with LBP, LPHD scheduling, and RD-RBP. The gain is modest, but it shows that probability-based scheduling benefits from message-level refinement.

cs.IT↗

arXiv2026

On the Intractability of the Minimum Distance Problem for Regular LDPC Codes

The minimum distance problem (MDP) for low-density parity-check (LDPC) codes is a central problem in coding theory and is closely related to the analysis of low-weight codewords and error-floor behavior. Although the unrestricted MDP is computationally intractable, its complexity under degree constraints that commonly occur in LDPC code design has remained less clear. In this paper, we study the MDP for left regular and biregular Tanner graphs. For every fixed $J\geq3$, we prove that the standard at-most-weight problem is $\mathrm{NP}$-complete for $J$-left regular Tanner graphs and that its exact-weight variant is $\mathrm{W}[1]$-complete when parameterized by the prescribed weight. For biregular Tanner graphs, we prove $\mathrm{NP}$-completeness for $(3,K)$-regular instances for every fixed $K\geq 3$ by replacing degree-two auxiliary completion blocks with a single-port high-girth gadget. A nonzero relative support inside this gadget induces an essentially cubic graph, so the Moore bound gives an exponential lower bound in the girth and allows a polynomial-size Karp reduction. Combining this right-degree amplification with a replica-and-global-check left-degree amplification yields $\mathrm{NP}$-completeness for $(J,K)$-regular Tanner graphs for every fixed $J,K\geq 3$. The reductions are based on a degree-preserving transformation framework consisting of hyperedge decomposition, check node splitting, and controlled variable replication. These transformations relate different degree distributions while preserving explicit maps among nonzero codewords, even covers, and nonempty $(a,0)$-trapping sets. The results delineate the computational limits of computing minimum distance exactly under natural regularity constraints.

cs.CC↗

arXiv2026

Budgeted Act-or-Defer Multi-Agent LLM Deliberation with Local Reliability Bounds

Multi-agent deliberation among LLMs can improve reasoning, but deployment requires deciding when the current answer is reliable enough to act on and when it should be escalated to human review. We formulate this as budgeted act-or-defer decision making. At each round, the system maps the debate prefix to a low-dimensional state, computes a $k$-nearest-neighbor lower confidence bound on state-conditional correctness using calibration data, and acts only when the bound exceeds a user-specified reliability threshold. The certificate controls wrong actions through the decomposition $β= δ+ α+ \varepsilon_{\mathrm{act}}$, separating calibration failure, residual action risk, and representation gap. The guarantee is conditional, not distribution-free: it relies on a valid local bias envelope and an action-region representation-gap bound, and each assumption is paired with falsification-style diagnostics. Because the same absolute wrong-action budget has different meanings across tasks of different difficulty, we set budgets relative to each task's final-round error using training data only, and evaluate safety by normalized budget usage $\mathrm{WA}/β$. On six benchmarks against nine baselines, the method uses 9--12% of the pre-declared budget on activated datasets, reaching up to 84% automation and 96% acted-on accuracy; on stress-test datasets, it defers rather than forcing unreliable automation. Rather than relying on per-task post-hoc threshold search, the method prospectively converts a user-declared wrong-action budget into an auditable act-or-defer operating point before deployment, under explicitly stated assumptions.

cs.AI↗

arXiv2026

Yuvion VL: A Multimodal Foundation Model for Adversarial Content and AI Safety

General-purpose models often struggle to reliably identify and understand real-world multimodal risks, largely due to the inherent multimodal adversarial nature of content and AI safety. We present Yuvion VL, a family of multimodal large language models purpose-built for content and AI safety, with both instruction-tuned and reasoning-oriented variants. Yuvion VL addresses this gap by treating safety as an inherently adversarial and multimodal problem and designing the entire pipeline around adversarial robustness. For data construction, we develop an automated pipeline integrating adversarial-aware data synthesis with multi-stage quality control, producing large-scale, high-quality multimodal samples augmented with domain knowledge and reasoning annotations. For training, we adopt a three-stage pipeline that includes continued pretraining for risk-concept cross-modal alignment, instruct post-training for production-grade safety tasks, and reasoning post-training for enhanced interpretability and performance in complex tasks. We further introduce Confuse-then-Contrast Fine-Tuning, a contrastive framework that mines model-specific confusions and constructs multi-image contrastive groups to enforce explicit discrimination of fine-grained visual-semantic elements, enabling the model to distinguish between visually similar cases with different safety implications in adversarial safety tasks. To support rigorous evaluation, we further introduce Yuvion VL RiskEval (YVRE), a collection of benchmarks covering diverse open and internal evaluations, with a focus on content and AI safety, adversarial robustness, and real-world capability requirements. Experiments show that Yuvion VL-32B achieves industry-leading safety performance, surpassing comparably sized open-source models and best closed-source commercial models, while maintaining comparable general capabilities.

cs.CV↗

arXiv2026

Yuvion LLM: An Adversarially-Aware Large Language Model for Content And AI Safety

As large language models are increasingly deployed in real-world systems, safety failures can still lead to harmful outputs and dangerous misuse. We argue that the essence of safety is adversarial: many failures arise not from natural inputs alone, but from strategic attempts to evade model policies and safeguards. However, existing general-purpose model development largely overlook this adversarial nature, and often remain insufficient for realistic safety scenarios involving planning, tool use, and multi-step reasoning, causing measured safety performance to overestimate real deployment robustness. To address this gap, we present Yuvion LLM, a large language model built for adversarially robust content safety and broader AI safety. Yuvion LLM treats adversarial robustness and agentic capability as first-class objectives. Its pipeline combines adversarially aware data construction, knowledge-enhanced continued pretraining, and policy-grounded multi-task safety post-training, including risk-aware supervised fine-tuning and reinforcement learning-based policy optimization, together with safety-aware agentic reinforcement learning for tool use and multi-step reasoning in complex safety scenarios. We further introduce the Yuvion LLM RiskEval (YLRE), a collection of 93 benchmarks across four evaluation categories, covering diverse open and internal evaluations with a focus on safety, adversarial robustness, and real-world capability requirements. Across these evaluations, Yuvion LLM demonstrates clear advantages on safety-focused benchmarks and particularly strong robustness under adversarial conditions, while maintaining solid overall capability. Notably, Yuvion-8B outperforms most state-of-the-art baselines, including substantially larger models such as GPT-5.4 and Qwen3-MAX, on several safety tasks.

cs.CL↗