30
30
Unlocking Lossless Speedups in LLMs via Discrete Diffusion
Uno pairs autoregressive LLMs with lightweight diffusion weights to draw multiple tokens in parallel, delivering up to 3x lossless speedup without a draft model.
The paper introduces diffusion-augmented LLMs: autoregressive weights trained with the standard next-token objective plus lightweight diffusion weights trained via a Diffusion Distillation phase to emit multiple tokens in parallel. Psi-Spec samplers enable lossless acceleration without the separate draft model required by speculative decoding. The 8B Uno model outperforms the 26B open DiffusionGemma and proprietary Mercury 2 on agentic tool use, coding, and long-context reasoning benchmarks, with up to 3x throughput gains over the base model at all evaluated batch sizes. Code and checkpoints are released publicly.
55
Nuha-Speech: Building General-Purpose Arabic Speech-LLMs
Nuha-Speech initiative builds general-purpose Arabic speech-LLMs using a 1.5M-sample speech QA corpus and fine-tuned Qwen-Omni variants.
The paper introduces Nuha-Speech, an initiative covering dataset construction, model training, and evaluation for Arabic speech large language models. The authors built an Arabic Speech Question-Answering corpus of over 1.5 million training samples and used it for supervised fine-tuning of Qwen-Omni model variants at multiple scales. A tailored evaluation framework with diverse tasks and metrics is designed to assess Arabic speech capabilities under limited resource constraints.
30
Retrofitting Code Using LLMs to Support Exceptional Behavior
EXCODER combines static/dynamic analysis with LLMs to retrofit exception-handling code, achieving 85.92% pass@1 with Qwen 2.5 Coder 32B on Java benchmarks.
The paper introduces the task of retrofitting existing code with Exception Related Code (throw statements, guarding conditions, try/catch blocks) so that given Exceptional Behavior Tests pass. EXCODER performs context engineering by integrating static and dynamic program analysis output with LLMs; it was evaluated on a benchmark built from 304 methods across 75 GitHub Java projects. Combined with Qwen 2.5 Coder 32B, EXCODER achieves pass@1, 5, and 10 rates of 85.92%, 86.18%, and 86.51%, roughly 13 percentage points over baseline, and manual inspection reveals remaining limitations.
25
CodeTD: Topology of Attention Detects Hallucinations in Code LLMs
CodeTD detects hallucinations in code LLMs before execution by analyzing topological patterns of attention maps, outperforming recent baselines.
CodeTD applies topological data analysis (TDA) to code LLM attention maps to quantify prompt-generation mismatch as a pre-execution correctness signal. Experiments cover HumanEval, MBPP, BigCodeBench, and MultiPL-E across 5 programming languages and 10 code LLMs up to 34B parameters. The method outperforms recent baselines and transfers between coding benchmarks, helping catch code that fails the task or embeds security vulnerabilities.
24
TempCloze: Can Video-LLMs Identify the Missing Middle?
TempCloze benchmark tests Video-LLMs' temporal reasoning with 1,521 videos, finding temporal alignment is the primary failure mode across 31 models.
TempCloze is a video cloze benchmark in which models must identify the true missing middle clip given the beginning and ending clips, using 1,521 carefully filtered videos from seven sources, mostly long-take and egocentric footage. Distractors are constructed along three dimensions: Semantic, Alignment and Progression, with shared scenes and objects to reduce appearance cues. Evaluation of 10 proprietary and 21 open-source Video-LLMs found Alignment is the primary bottleneck, with models often recognizing plausible semantics and local event progression but struggling with temporal alignment.
25
I wrote an AI textbook — how long until AI can do it better?
AI researcher Nathan Lambert argues LLMs remain weak at long-form technical writing, questioning whether models can autonomously organize scientific knowledge for breakthroughs.
Nathan Lambert describes writing a post-training textbook, Reinforcement Learning from Human Feedback, and finds today's LLMs weak at organizing long-form technical content despite becoming superhuman at coding and math. He notes GPT 5.5 Pro found deep typos across a 200-300 page manuscript while Claude models proved more useful as editors. He argues that compressing knowledge through writing is a prerequisite for autonomous scientific insight and tempers expectations for near-term AI-driven open science.
24
How much of F-Droid is LLM generated?
A FOSS maintainer manually graded 102 F-Droid apps from the September 12, 2026 update batch, finding many show signs of LLM-generated code.
A student and FOSS app maintainer reviewed 102 apps pushed to F-Droid on September 12, 2026, assigning each a three-tier rating for likelihood of LLM-authored code (mostly AI >50%, hard to say/mostly human, no signs of AI). The heuristic relies on commit aesthetics, README and branding style, and the presence of agentic infrastructure like Claude Code or Codex, which automatically places an app in the 'mostly AI' tier. Example ratings include Amber (Nostr event signer) as mostly AI, and Aria for Misskey as showing no AI signs. The author stresses reliable detection of LLM-generated code from text alone is impossible, so ratings are approximate.
15
Before You Poll with LLMs: A Deliberative Diagnostic Framework
Deliberative diagnostic shows all five tested frontier LLMs misrepresent human belief shifts after arguments, with GPT-5.1 reversing on outgroup questions.
The Deliberative Polling Diagnostic Framework compares human and LLM persona belief shifts after identical informational interventions, using data from America in One Room (526 personas, 72 questions). All five frontier models tested failed uniquely: GPT-5.1 exhibited partisan reversal (80% on outgroup vs 26% on policy questions), Gemini 2.0 Flash, Claude Sonnet 4.5 and Llama 3.3 70B overshot at 5-7x human magnitude, and DeepSeek V3 showed near-zero change (rigidity). The authors term the underlying signature 'self-sycophancy', conformity to the model's internal persona stereotype rather than reasoning from provided information.
40
xDailyBench: Benchmarking LLMs on Professional Consultation for Real-Life Problems
xDailyBench tests 11 frontier LLMs on 248 real-life consultation tasks; the best models score 75.6% and lag on implicit requirements.
The benchmark spans 51 scenarios across personal life, white-collar work, learning and research, and cross-domain activities, grounded in requests users actually completed or intended to complete with AI. Tasks are scored with fine-grained binary rubrics covering explicit and implicit requirements under standardized agentic settings. Across 11 frontier models, the best achieved a 75.6% task-level score, with all models performing at least 9 percentage points worse on implicit than explicit requirements.
26
30
Large Language Models Develop Belief State Geometry In-Context
Probing six open-source LLMs on HMM-generated data shows belief states are linearly decodable (R² 0.83–0.99), suggesting in-context learning approximates Bayesian prediction.
Researchers prompted six open-source LLMs with data from 40 hidden Markov models selected for non-trivial belief structure and probed residual-stream activations for belief states (posteriors over hidden states). Belief states were linearly decodable with peak R² values of 0.83–0.99 across HMM/LLM combinations, spanning early to late layers. Patching and steering the probe-identified subspace preserved downstream prediction quality while control interventions degraded performance substantially, establishing functional relevance. The results provide representation-level evidence that in-context learning approximates optimal Bayesian prediction over a context-inferred generative model.
45
Learning to Solve Hard Problems in RL for LLMs by Never Giving Up
Paper introduces Never Give Up adaptive sampling, fixing RL's 'Matthew Effect' where compute is wasted on easy problems and hard problems see little improvement.
Researchers identify a 'Matthew Effect' in reinforcement learning for LLMs, where RL yields large gains on easy problems but minimal improvement on hard ones because compute is misallocated. They propose Never Give Up (NGU), an adaptive sampling method that keeps generating samples for a problem until one is correct, using asynchronous RL to filter easy problems cheaply and concentrate compute on hard ones. NGU improves performance per compute on the Deepscaler math benchmark and iteratively solves the Manufactoria coding task where standard GRPO with per-test reward fails.
45
Why Is Video Still So Expensive? A Survey of Inference-Efficiency Mechanisms in Video and Audiovisual LLMs
A survey catalogs inference-efficiency techniques for video and audiovisual LLMs, mapping bottlenecks in sampling, encoding, token reduction, and LLM decoding.
This survey covers inference-efficiency mechanisms for visual and audiovisual video LLMs, reporting reductions in parameters, FLOPs, latency, memory, and token counts. It organizes methods by pipeline stage, covering frame sampling, modality encoding, connector-level token reduction, and LLM prefilling and decoding for systems built since late 2022. The authors compile accuracy-cost comparisons under shared host models and input protocols, identify gaps in audiovisual efficiency and standardized evaluation, and maintain a public repository.
40
ReactHuman: A Physics-Grounded Benchmark for Human-Like Reactive Decision-Making in Embodied Multimodal LLMs
ReactHuman benchmark tests whether multimodal LLMs react safely to sudden household hazards; seven evaluated models mishandle roughly one hazard in three.
ReactHuman is the first physics-grounded benchmark for human-like reactive decision-making, placing a multimodal LLM as the brain of a simulated humanoid facing 17 event families of sudden household hazards across over 1,000 bit-for-bit reproducible scenes with annotation-free ground truth from 240 Hz rigid-body simulation, including adversarial objects whose appearance contradicts their physics. A five-metric suite scores each reaction along reasonable, safe, and physically grounded axes, and every committed plan is physically executed. Seven representative MLLMs mishandle roughly one hazard in three, act from fixed dispositions rather than the observed scene, trust appearance over motion, and miss interception points at meter scale; none of these failures shrink with model scale.
42
SAFIRE: Safety-Critical Benchmark for Fine-grained Fire and Smoke Understanding in Multimodal LLMs
SAFIRE, an 83K-image fire and smoke benchmark, shows open-source multimodal LLMs average only 61.9% accuracy on safety-critical fire reasoning.
SAFIRE is a large-scale benchmark for fire-smoke understanding in multimodal LLMs with 83K captioned images across 20 scenarios and 193K multiple-choice VQA questions spanning 10 evaluation dimensions from perception to higher-order reasoning. Annotations were built via a GPT-5.4-assisted multi-stage pipeline with MLLM majority voting. Ten open-source MLLMs (8B-38B) average 61.9% accuracy, exposing major gaps in safety-critical reasoning. Adapting vision encoders on 7% of the domain data raises fire-scene classification from 20.1% to 64.5%.
33
You Can't Prefer Emotions You Don't Sample: Intensity Undershoot in DPO-Tuned LLMs
Study quantifies DPO-tuned LLMs undershooting requested emotional intensity, tracing the gap to candidate-pool extremity rather than conditioning format.
Conditioning an instruction-tuned LLM on continuous valence-arousal targets yields gain of only 0.26 for valence and 0.13 for arousal on Llama-3.1-8B, far below faithful control of 1.0. The authors attribute undershoot to neutral-heavy preference corpora like EmoBank and candidate pools lacking extreme affect, leaving DPO without extreme exemplars. Uniform target coverage with a hotter candidate pool raises valence gain to 0.40 on Llama-3.1-8B and 0.44 on Qwen3-8B, with modest in-distribution cost; arousal gains remain unstable across seeds.
31
The AI ‘Ghosts’ Contaminating Academic Publishing
Samsung and University of Warsaw researchers find LLMs repeatedly generate the same fake author names, contaminating academic records with 1,655 ghost-authored DOIs.
A preprint from Samsung and the University of Warsaw, "The Ghost Couple: Correlated LLM Name Priors and Their Haunting of the Web and Academic Publishing," shows that LLMs such as Claude, ChatGPT, and Gemini repeatedly generate the same fictional names like Elena Vasquez, Marcus Chen, and Aris Thorne as experts and co-authors. Researchers identified 1,655 ghost-authored records on CERN-operated Zenodo carrying real DataCite DOIs, fabricated journals, and backdated publication dates. Ghost names also form synthetic research groups on ResearchGate and are indexed without verification by Google Scholar and Semantic Scholar. The researchers suggest correlated name priors could serve as provenance signals for detecting AI-generated content.
32
Can LLMs Engineer Their Own Agent Harness? ByteDance Seed’s HarnessDev Says Only 34 of 64 Changes Generalize
ByteDance Seed's HarnessDev benchmark finds LLM-built agent harnesses trail human engineering on code and search, with only 34 of 64 revisions generalizing.
Researchers from ByteDance Seed, SUTD, Georgia Tech, M-A-P, and TokenWave.AI introduce HarnessDev, a benchmark that evaluates the runnable agent harness an LLM writes rather than its answers, using Creation and Evolution stages across SWE-bench Pro, Terminal-Bench 2.1, MLE-bench, EQ-Bench3, and BrowseComp (2,207 instances). Six creator models including Opus 4.8, GPT-5.5, Gemini 3.1 Pro, DeepSeek V4 Pro, Qwen 3.7 Max, and Seed 2.0 Pro were tested; Opus 4.8 posted the best average of 67.8 versus an 86.2 human-engineered reference. Self-built harnesses beat references on writing and ML experimentation but lag badly on code and search, and quality proved executor-specific: Opus 4.8 fell from 69.3 to 33.0 on SWE-bench Pro when the executor was switched to Gemini. Evolution gains were small and noisy: of 64 adjacent changes, feedback and held-out scores agreed only 34 times (53.1%), and much generated state and memory code never executed.
48
RetroThinker: Enabling Retrospective Thinking in Speech LLMs
RetroThinker is a post-training framework letting the Moshi speech LLM self-correct reasoning mid-stream, adding 11% GSM8K accuracy at similar latency.
Researchers introduce RetroThinker, a multi-stage post-training framework that equips the Moshi speech LLM to verify and forward-correct chain-of-thought steps during streaming inference. It combines supervised fine-tuning on curated retrospective thinking data with length-based direct preference optimization (DPO). On GSM8K it achieves an 11% absolute accuracy gain over non-retrospective baselines at comparable latency.
28
Forgetting Only What Matters: Layer-Selective Unlearning toward Robust LLMs
Researchers introduce FOM-UL, a layer-selective machine unlearning framework that improves forgetting-utility trade-offs and resists knowledge recovery after quantization.
FOM-UL selects transformer layers for unlearning using a forget-to-retain significance score, concentrating parameter updates on layers highly influential for the forget set while leaving most of the model unchanged. It reduces residual memorization versus GA, NPO, KLD, SURE, ReLearn, and LUNAR-based baselines on TOFU, KnowUnDo, and MUSE-style evaluations while preserving retain-set utility. Under 8-bit and 4-bit post-training quantization and adversarial prompts, it maintains stronger suppression of forgotten content, addressing brittleness of diffuse unlearning updates.
36
Φ-Bench: Can Large Language Models Engineer the Infrastructure That Powers Them?
Researchers release Phi-Bench, a benchmark evaluating frontier LLMs on open-ended, long-horizon engineering and optimization of the LLM infrastructure stack.
Phi-Bench evaluates LLMs on open-ended engineering of the LLM infrastructure stack, derived from optimization problems studied in frontier research and grounded in real-world code repositories. Tasks range from localized kernel-level function completion to long-horizon implementation and end-to-end system optimization. Experiments on frontier LLMs reveal current capabilities and limitations on the path toward autonomous optimization of future AI infrastructure.
28
Breaking the 1.58-bit Barrier for Ternary LLMs
An arXiv paper claims a method that breaks the 1.58-bit barrier for ternary large language models.
The arXiv preprint 2609.16338, titled 'Breaking the 1.58-bit Barrier for Ternary LLMs,' presents research on ternary-weight large language models, which use roughly 1.58 bits per weight. The source text contained only the title and Hacker News engagement data (56 points, no comments), so further technical details are not available.
35
When Should LLMs Abstain? Chain-of-Self-Questioning for Selective Risk Control
Chain-of-Self-Questioning prompting cuts LLM wrong-answer commitments 32% relative while raising answered accuracy, holding across eleven model families.
The paper introduces Chain-of-Self-Questioning (CoSQ), a prompt-only framework that makes LLM answer commitment conditional on an explicit assessment of the information required to answer. On an 817-item TruthfulQA multiple-choice set, Grounded-CoSQ at τ=0.90 reduced mean unconditional wrong-commitment rate from 13.1% under chain-of-thought to 8.9% (a 32.1% relative reduction), while raising answered accuracy from 86.9% to 89.7% at 87.6% coverage. Improvements held across eleven open-weight and hosted model families and at every evaluated threshold, with convergent evidence from a Natural Questions short-answer evaluation.
30
Enhancing Accessibility of Medical Texts through Large Language Model-Driven Plain Language Adaptation
Study shows LLMs with Mixture-of-Agents and QLoRA finetuning effectively simplify medical texts into plain language while preserving content.
The paper evaluates Plain Language Adaptation (PLA) using GPT-4o-mini, Gemini-1.5-pro, and LLaMA in zero-shot and few-shot settings. It compares prompting strategies, QLoRA finetuning across models, and integrates Mixture-of-Agents (MoA) techniques for robustness. Results demonstrate LLM-driven PLA makes healthcare texts more comprehensible while retaining essential content.
15
Verifiable by Construction: Claim-Level Evaluation of Verbatim Citation in Clinical Question Answering
Evaluation of twelve LLMs on 222 clinical questions shows verbatim quotes rarely substantiate claims; claude-opus-5 fully substantiates only 37.1%.
The authors build a standardized harness over four clinical practice guidelines and evaluate twelve LLMs on 222 synthetic clinical questions, measuring citation attachment, verbatim quote production, and claim substantiation. Most models attach verbatim quotes to over 90% of claims from prompting alone, though lightweight models like claude-haiku-4.5 struggle. Quotes frequently fail to substantiate claims: claude-opus-5 quotes 98.0% of claims but fully substantiates only 37.1%, exposing a capability gap for verifiable clinical QA.
38
HypoEvolve: Genetic Algorithms Enable Multi-Agent LLMs to Discover Scientific Hypotheses
HypoEvolve couples a generational genetic algorithm with specialized LLM agents to generate drug-repurposing hypotheses, beating six baselines on DepMap selectivity (0.171 vs 0.115).
HypoEvolve coordinates specialized LLM agents through a generational genetic algorithm in which scientific judgments and new proposals reshape a hypothesis population. Evaluation centers on drug repurposing, linking mechanistic explanations to target-level biological claims assessed via external measures adapted from DepMap and Open Targets. Across 34 cancer types, HypoEvolve scores highest against six baselines on both measures, with DepMap selectivity of 0.171 versus 0.115 for the strongest baseline, and gains generalize to held-out cancer types.
28
HypoEvolve: Genetic Algorithms Enable Multi-Agent LLMs to Discover Scientific Hypotheses
HypoEvolve uses a generational genetic algorithm coordinating specialized LLM agents to generate scientific hypotheses, outperforming six baselines on cancer drug repurposing.
HypoEvolve is a framework that coordinates specialized LLM agents through a generational genetic algorithm to produce, revise, and retain scientific hypotheses with explicit collaboration roles. It evaluates drug repurposing hypotheses against external evidence from DepMap and Open Targets across 34 cancer types. It achieves the highest scores against six baselines, reaching DepMap selectivity of 0.171 versus 0.115 for the strongest baseline, and gains over single-pass generation generalize to held-out cancer types.
40
From Parameters to Answers: How LLMs Retrieve and Use Their Internal Knowledge
Interpretability study traces how Qwen, Llama, and Gemma route query information and internal knowledge across layers when answering questions.
Researchers used layerwise interventions on hidden states to separate query-routing signals from target knowledge in language models, testing Qwen, Llama, and Gemma on country-continent questions with varied answer types. A pair-conditioned request direction strengthens before interventions alter downstream knowledge, opening a causal window while answer-supporting content is still forming. Trajectories differ by model: Gemma shows a partially overlapping mid-layer routing profile, while Llama has no sustained routing-effect window under the same gates.
25
X-AuT: Progressive Audio-Encoder Compression for Speech LLMs with Cross-Scale Distillation
XPeng AI's X-AuT prunes speech LLM audio encoders, cutting Qwen3-ASR-0.6B error from 5.61% to 5.27% with fewer parameters.
X-AuT is a progressive compression framework for speech LLM audio encoders that selects layer combinations via short behavioral probes and restores pruned models using cross-scale distillation and LoRA finetuning while keeping the language-model backbone frozen. Compressing Qwen3-ASR-0.6B from 18 to 16 audio-encoder layers lowered macro-average error from 5.61% to 5.27% on ten Chinese-English benchmarks. A 14-layer model reached 5.75% error with 20.7% fewer audio-tower parameters, and progressive pruning outperformed direct pruning (5.75% vs 6.73%).
26
Fortunate Recall: Ontology-Driven Memory Lifecycle Management for Persistent Coherence in LLMs
Fortunate Recall introduces ontology-based lifecycle policies for LLM memory, cutting confabulation roughly in half (e.g., 45.1% to 22.4%) versus Mem0.
Fortunate Recall (FR) is a composable policy layer that classifies personal facts into a 10+1 behavioral ontology and applies category-specific lifecycle rules including differential temporal decay, slot-key supersession, event-time validity, and retrieval routing. FR-Bank scores 76.9% on the new 516-question LifecycleBench, ahead of Mem0, A-MEM, Memory-R1, and MemoryOS (61%-70.5%), and 75.2% on LongMemEval-S. End-to-end, confabulation drops from Mem0's 45.1% to 22.4% over answered queries, with the ranking replicating on open-weight Kimi K2.5 and transferring to the independent BEAM benchmark (46.8% vs 32.9%).
30
Reference-Based Bias Detection in LLMs via Relative Representations of Hidden States
Researchers propose auditing LLM bias via relative hidden-state representations, detecting bias increases with 3-50x less compute than output-level benchmarks.
The paper introduces a reference-based bias auditing method that compares hidden-state representations across model variants, such as before and after fine-tuning, by encoding sentences relative to a fixed anchor set. The resulting Representational Bias Shift (Delta-B) correlates with output-level bias change in 15 of 18 tested settings, reaching |r| = 0.84 under full fine-tuning across WildGuardMix, DecodingTrust, and ToxiGen benchmarks. Thresholding Delta-B detects checkpoints whose bias increased with ROC AUC between 0.65 and 0.99 and beats a SEAT-based baseline, while auditing a model in about three minutes with 3-50x less compute.
42
RESCUE-BENCH: Towards Relation-Aware Multi-Party Emotional Support Conversation Systems
Researchers introduce RESCUE-Bench, a video benchmark of 191 couple and family conversations evaluating LLMs on relation-aware multi-party emotional support.
RESCUE-Bench is built from real couple and family interview conversations, containing 191 samples, 7,079 annotated turns, and 1,064.8 minutes of video. It defines six tasks measuring two capabilities: Relational Understanding and Relation-Sensitive Support. Experiments with ten LLMs show models handle local emotional cues but struggle with relation pattern prediction, viewpoint prediction, and support strategy prediction.
27
Beneath the Surface of Chains-of-Thought: A Mechanistic Interpretation of Reasoning Operations in LLMs
Study shows LLM reasoning operations like planning and deduction are geometrically separable in hidden states, with separability peaking in middle layers.
Researchers investigate whether functional reasoning operations — problem formulation, goal decomposition, deduction — have corresponding geometric structure in LLM hidden representations. They find operations are separable in held-out representations with separability peaking in middle layers, ruling out lexical and positional confounds; token-wise operation alignment becomes more distributed across layers, and identical surface tokens are represented differently depending on their surrounding chunk. Attention-masking interventions show chunk-onset operation-aligned representations depend on preceding reasoning context; code is released on GitHub (naver-ai/beneath-cot).
40
What Else Needs Fixing? Exploring Cost-Effective Test-Time Compute for Revision Propagation in Artifacts Generated Through Conversation
A new benchmark shows LLMs reach 68.3-93% accuracy propagating local revisions across conversationally generated artifacts, with parallel-sample selection most cost-effective.
The paper introduces a benchmark for revision propagation: when users request a local change, LLMs must identify dependencies and update all affected parts of an artifact generated through conversation, where context lives in the chat history. Nine revision methods, including sequential reflection and parallel sampling variants, were evaluated on gpt-oss-20b/120b, gpt-5.4-mini, and qwen3.5-9b/27b/122b. Baselines scored 68.3-93% accuracy, and selecting among three parallel samples via LLM-based or medoid selection improved accuracy by 2.2-9.7% as the most cost-effective test-time compute strategy. Code and dataset are released.
30
Revisiting Complete Reasoning Traces for Post-Training
Researchers show full reasoning traces provide limited benefit in LLM post-training, with heavily truncated or endpoint-only trajectories performing comparably.
A pilot study plus attention-based analyses and controlled token-removal studies show intermediate tokens in reasoning trajectories contribute minimally to final reasoning quality. Partial trajectories remain effective even under heavy truncation, and training on endpoints alone leads to consistent changes in reasoning behavior. The finding also benefits reinforcement-learning and on-policy distillation post-training; code is released at github.com/naver-ai/revisiting-trace.
32
Large Language Models for HVAC Operations in Building Energy Systems: A Critical Review of Methods, Applications, and Deployment Readiness
Systematic review of 66 studies finds LLMs for HVAC operations are mostly research-stage, with no ready-now deployment and only four pilot-level studies.
A critical review of 66 peer-reviewed studies from 2023 to March 2026 examines LLMs for HVAC operations in building energy systems. Only four studies reach pilot-level evidence, none reports sustained operational deployment, and 63 of 66 are research-only. Conventional ML, MPC, and RL remain dominant for high-frequency control and short-horizon forecasting, and the evidence supports LLMs primarily as semantic and workflow layers rather than autonomous controllers.
45
When Models Edit Too Much: On the Fidelity of Minimal Code Edits
A 400-task BigCodeBench evaluation shows frontier LLMs widely over-edit code; a preservation instruction cuts excess edits and raises Pass@1 by 2.3 points.
Researchers built an evaluation framework from 400 BigCodeBench problems with injected AST-level corruptions, each with a known minimal patch, to measure over-editing in LLM code repair. Even strong models like GPT-5.5 produce unnecessarily large edits despite high Pass@1. Adding a preservation instruction reduced average excess Levenshtein distance from 0.195 to 0.131, cut added cognitive complexity by 26.6%, and raised Pass@1 by 2.3 points. Reinforcement learning post-training gave the best out-of-domain edit-fidelity trade-off, while supervised fine-tuning overfit to seen corruption patterns.
30
A Zeroth-Order Paradigm for LLM Preference Alignment
Researchers propose ComPO, a zeroth-order comparison-based preference alignment method with convergence guarantees that mitigates likelihood displacement in LLMs.
ComPO extracts directional information from preference pairs via comparison oracles instead of optimizing a differentiable preference loss, addressing likelihood displacement in direct alignment methods. The paper establishes convergence guarantees for the offline scheme and introduces an online variant with reverse-KL control using unlabeled policy generations. Experiments on Mistral, Llama, Gemma-2, Gemma-3, and Qwen3 models show improvements over existing direct alignment methods, including length-controlled win rates.
32