arXiv subjects

Frederic Sala

Publications and source records attributed to Frederic Sala.

At least 19 recordsLinked to original sources

arXiv2026

Can Generalist Agents Automate Data Curation?

Curating training data is among the most consequential yet labor-intensive parts of modern AI development: practitioners iteratively propose, implement, evaluate, and revise data policies against noisy benchmark feedback. We ask whether generalist coding agents can automate this data-curation loop. We introduce *Curation-Bench*, an agent-centric benchmark that fixes the model, training recipe, and evaluation suite while giving agents command-line access to inspect data, implement policies, submit them to a fixed training/evaluation pipeline, and revise. In a vision-language instruction-tuning instantiation, out-of-the-box agents reach strong published data-selection baselines within ten iterations. However, trajectory analysis reveals a persistent *execution-research gap*: agents mainly tune local policy variants rather than explore new policy families, even when given strategy guides and paper references. Scaffolds requiring each iteration to cite, instantiate, and adapt a prior method shift agents toward method-guided exploration. The scaffolded agent autonomously composes -- without human design input -- a data-selection policy that outperforms strong published baselines at one-tenth their data budget. Overall, current agents can run the curation loop, but reliable data research requires scaffolded method adaptation, not open-ended prompting alone. Code and benchmark are open-sourced.

cs.AI↗

arXiv2026

MAGS: Multi-agent Auto-formalization Guarantees Safety for Agentic Outputs

LLM coding agents now generate complex programs at a scale that makes thorough human review increasingly difficult, raising the risk of safety and security failures. Common approaches, including fuzz testing, static analysis, and LLM-as-a-Verifier, can detect many failures but struggle to cover all possible edge cases. Formal verification addresses this by providing machine-checkable guarantees over specified properties, but traditionally demands substantial manual specification and proof engineering. We introduce a unified multi-agent framework, MAGS, that generates executable programs with formal safety guarantees, using Dafny as a verification-aware intermediate representation where safety properties can be mechanically checked. MAGS formalizes and freezes human-audited APIs and safety requirements, translates generated code into Dafny, repairs violations using verifier feedback, and compiles verified programs back into executable code. We evaluate MAGS on 100 CUDA kernels, 100 terminal scripts, and 20 robotic-arm tasks. Across all 220 examples, it achieves a 100% success rate in producing programs with non-trivial safety guarantees against frozen specifications. Independent safety and functional evaluations further show strong performance across all three domains, while revealing failures when the auto-formalized semantics do not fully capture the target behavior.

cs.AI↗

arXiv2026

Register Tokens for Bounded-State Reasoning in Diffusion Language Models

Masked diffusion language models (dLLMs) generate text by iteratively denoising masked tokens with bidirectional attention. Extending reasoning across generation chunks normally requires keeping earlier generated text in context. We ask whether a dLLM can instead continue reasoning after that text is cleared, using only a fixed-size carried state. We implement this state as a small number of register tokens: dedicated fixed-position tokens whose continuous hidden states are trained to carry reasoning progress across generation chunks. We post-train dLLMs to decode a chunk of text, clear it while preserving the register values, and continue decoding from the prompt and carried state. In our main comparisons on LLaDA and Dream, registers outperform discrete-text carry on every benchmark, with gains of up to 8.5 points on math and 19.5 points on code. Registers are especially effective for bounded code generation, where correct programs usually span several chunks. Finally, registers can be further refined with reinforcement learning on long-horizon reasoning tasks.

cs.CL↗

arXiv2026

CARE: Confounder-Aware Aggregation for Reliable LLM Evaluation

LLM-as-a-judge ensembles are the standard paradigm for scalable evaluation, but their aggregation mechanisms suffer from a fundamental flaw: they implicitly assume that judges provide independent estimates of true quality. However, in practice, LLM judges exhibit correlated errors caused by shared latent confounders -- such as verbosity, stylistic preferences, or training artifacts -- causing standard aggregation rules like majority vote or averaging to provide little gain or even amplify systematic mistakes. To address this, we introduce CARE, a confounder-aware aggregation framework that explicitly models LLM judge scores as arising from both a latent true-quality signal and shared confounding factors. Rather than heuristically re-weighting judges, CARE separates quality from confounders without access to ground-truth labels. We provide theoretical guarantees for identifiability and finite-sample recovery under shared confounders, and we quantify the systematic bias incurred when aggregation models omit confounding latent factors. Across 12 public benchmarks spanning continuous scoring, binary classification, and pairwise preference settings, CARE improves aggregation accuracy, reducing error by up to 26.8\%. Code is released in \href{https://github.com/SprocketLab/CARE}{https://github.com/SprocketLab/CARE}.

cs.LG↗

arXiv2026

Shrinking the Generation-Verification Gap with Weak Verifiers

Verifiers can improve language model capabilities by scoring and ranking responses from generated candidates. Currently, high-quality verifiers are either unscalable (e.g., humans) or limited in utility (e.g., tools like Lean). While LM judges and reward models have become broadly useful as general-purpose verifiers, a significant performance gap remains between them and oracle verifiers (verifiers with perfect accuracy). To help close this gap, we introduce Weaver, a framework for designing a strong verifier by combining multiple weak, imperfect verifiers. We find weighted ensembles of verifiers, which typically require learning from labeled data, significantly outperform unweighted combinations due to differences in verifier accuracies. To reduce dependency on labeled data, Weaver leverages weak supervision to estimate each verifier's accuracy and combines outputs into a unified score that better reflects true response quality. However, directly applying weak supervision algorithms poses challenges, including inconsistent verifier output formats and handling low-quality verifiers. Weaver addresses these using dataset statistics to normalize outputs and filter specific verifiers. We study Weaver's effectiveness in test-time repeated sampling, where a model generates multiple candidate responses and selects one. Our evaluations show Weaver significantly improves over Pass@1-performance when selecting the first candidate-across reasoning and math tasks, achieving o3-mini-level accuracy with Llama 3.3 70B Instruct as generator, and an ensemble of 70B or smaller judge and reward models as verifiers (87.7% average). This gain mirrors the jump between GPT-4o and o3-mini (69.0% vs. 86.7%), which required extensive finetuning and post-training. To reduce computational costs of verifier ensembles, we train a 400M cross-encoder using Weaver's combined output scores.

cs.CL↗

arXiv2026

Geospatial Diffusion-based Evolution Synthesis (GeoDES) for Storm-Centered Weather Augmentation

While machine learning-based weather models hold significant promise, they struggle to predict the detailed structure of large-scale weather systems such as cyclonic storms. Regional models are constrained by limited historical records within fixed geographic boundaries, while global models are computationally expensive and often operate at resolutions too coarse to capture fine-grained storm dynamics. To bridge this gap, we introduce the Geospatial Diffusion-based Evolution Synthesis (GeoDES) model, a custom image-to-video diffusion model. By focusing generation strictly on the evolving storm structure, GeoDES synthesizes physically consistent, high-fidelity weather events suitable for stress-testing forecast models and expanding meteorological datasets. Evaluations demonstrate that GeoDES outperforms prior methods on key metrics, achieving $52\%$ lower Peak Vorticity Error and $8\%$ higher Anomaly Correlation Coefficient than the next strongest methods on the North Atlantic test set.

cs.LG↗

arXiv2026

Models Can Model, But Can't Bind: Structured Grounding in Text-to-Optimization

Text-to-optimization requires two separable capabilities: modeling -- choosing the right optimization structure -- and binding -- grounding every coefficient, index, and parameter in the concrete problem data. We study this via Text2Opt-Bench, a scalable benchmark of solver-verified optimization problems spanning 12 categories, from textbook linear programs to stochastic and multi-objective formulations with up to thousands of variables. Across 10+ models, we find that accuracy collapses as instance data grows, even when the formulation itself is simple. We call this the effective binding limit. We study it with a family of techniques, BIND, that externalize numeric data to structured files so the model binds data programmatically rather than transcribing from the prompt. When using an oracle for externalizing data, we recover between 12 and 27 accuracy points, confirming binding as a key -- but recoverable -- failure mode. In a deployable setting without oracle access, we validate our hypothesis by finetuning a model exclusively on binding and show that it outperforms end-to-end SFT and RL across three structurally distinct optimization categories, with a 1.5B binding specialist alone matching a 7B end-to-end baseline.

cs.LG↗

arXiv2026

OSWorld 2.0: Benchmarking Computer Use Agents on Long-Horizon Real-World Tasks

Existing computer-use benchmarks fail to capture the realism, complexity, and long-horizon demands of real-world computer use, limiting their ability to reveal the limitations of frontier agents. We introduce OSWorld 2.0, a benchmark of 108 long-horizon computer-use workflows across everyday and professional tasks, designed to capture complex and challenging real-world phenomena. Each task represents a realistic end-to-end workflow that takes human users a median of about 1.6 hours to complete and requires an average of 318 tool calls with Claude Opus 4.7 using maximum thinking, compared with about 30 in OSWorld 1.0. OSWorld 2.0 targets challenge phenomena that are common in real workflows yet underrepresented in prior benchmarks, spanning interaction-design challenges such as streaming interaction and dynamic environments, as well as agent-pattern challenges such as cross-source reasoning, implicit-state inference, and visual-spatial precision. Tasks are grounded in authentic input artifacts and cross-referenced against realistic stateful user profile data, and include separate safety reports auditing safety-sensitive execution. Under our primary binary-completion metric at 500 steps, Claude Opus 4.8 with maximum thinking and batched tool calls scores best but still completes only 20.6% of tasks at a 54.8% partial score; GPT-5.5 is far more token-efficient yet plateaus near 13%. These results show that current agents are still far from professional-level computer use: rather than stumbling on basic GUI control or coding, they lose track of constraints, miss information that arrives mid-task, guess rather than ask the user, and skip verification, struggling most when a task hinges on hidden state they must recover.

cs.AI↗

arXiv2026

WARP: Weight-Space Analysis for Recovering Training Data Portfolios

Foundation models are routinely released to the public, yet the data recipes used to train them -- such as domain mixture weights that determine how different sources are sampled -- are rarely disclosed. This creates an access asymmetry: researchers study the resulting models but lack visibility into the training distribution that produces them. Prior works for inferring training data, such as membership inference, detect at the level of individual samples and thus cannot characterize the global composition of the training corpus. We introduce WARP, a framework that recovers a fine-tuned model's training mixtures directly from its released weights. WARP interpolates between the base and fine-tuned models using model merging, generating pseudo-checkpoints that approximate the missing training trajectory and expose a geometric footprint of the training data in the weight space. From these simulated footprints, WARP extracts geometric features and maps them to domain proportions using either a parameter-free softmax readout or an MLP projector trained on synthetic mixtures. In controlled experiments with BERT and GPT-2, WARP recovers domain mixtures with an average MAE as low as 0.046 and 0.104 respectively, outperforming membership inference and a variant with access to the true training trajectory.

cs.LG↗

arXiv2026

Counterexample Guided Learning in the Large using Reasoning Agents

LLMs and LLM agents should improve when given feedback, but identifying when they are able to do so is difficult: feedback is heterogeneous, domain-specific, and difficult to control. We approach this challenge by asking LLMs to perform regular-expression induction, a classical symbolic learning problem where precise mechanisms for feedback exist in the form of counterexamples. In counterexample-guided learning, a learner (LLM) proposes candidate regular expressions from positive/negative-labeled strings, and the teacher (verifier) returns counterexamples showcasing the difference between the candidate and target languages. We identify novel counterexample-guided refinement strategies that enable effective regex learning, such as regularization and symbolic counterexample clusters. We also explore agentic strategies such as reflection and repair loops. Empirically, we find that verifier feedback substantially improves sample efficiency on challenging regex-induction tasks, reducing the number of labeled examples required and enabling learning of complex target expressions where standard prompting fails. For example, on the hardest task groups, our counterexample-guided framework improves success from 3.2% to 38.1% and from 38.9% to 74.1% on two different regex domains. These results suggest that LLMs can benefit from rich feedback beyond treating it as additional data, opening the door for robust verifier-guided methods for LLM-based program synthesis and formal reasoning.

cs.LG↗

arXiv2026

Continual Learning Bench: Evaluating Frontier AI Systems in Real-World Stateful Environments

Continual learning, the ability of AI systems to improve through sequential experience, has attracted substantial interest, but no high-quality benchmark exists to evaluate it. We introduce Continual Learning Bench (CL-Bench), the first difficult, expert-validated benchmark designed to measure whether LLM-based systems genuinely improve with experience. CL-Bench spans six diverse domains (software engineering, signal processing, disease outbreak forecasting, database querying, strategic game-playing, and demand forecasting), each validated by domain experts and designed so that tasks share a learnable latent structure (codebase layout, disease outbreak dynamics, opponent strategies) that a stateful system can discover online but a stateless one cannot. We evaluate frontier models across several agent architectures, from naive in-context learning (ICL) to dedicated memory systems, introducing a gain metric to isolate learning from prior capabilities. We find that these systems leave headroom for improved continual learning: agents frequently overfit to immediate observations or fail to reuse knowledge across instances, and dedicated memory systems do not fix this -- in fact, naive ICL outperforms systems dedicated to memory management. CL-Bench is the first benchmark to evaluate continual learning across diverse real-world domains with expert-validated tasks and isolate online learning from underlying model capability, showing a need for better continual learning systems.

cs.AI↗

arXiv2026

Codifying the Judge: Scalable Evaluation via Program Distillation

LLM-as-a-judge has become the standard for automated evaluation, but it suffers from high cost, significant latency, and opaque decisions -- limitations that undermine its scalability and reliability. We address these with a simple, efficient alternative: program distillation. Instead of prompting an LLM at the evaluation time, we distill its decision logic into a committee of programs that score candidates directly. These programmatic judges offer transparency, are easily inspected or edited, and eliminate per-sample API costs. Building on this notion, we introduce PAJAMA, a system that synthesizes programs as judges, aggregates their decisions into a joint verdict, and incorporates a fallback mechanism to selectively escalate low-confidence cases to an LLM. Across five datasets and four model families, we show that programmatic judges can match the performance of a 13B-size LLM judge. When using program outputs as routing signals, PAJAMA improves both accuracy and throughput and advances the Pareto frontier. Beyond evaluation, programmatic judges produce cheap and effective reward signals: on RewardBench, a reward model distilled from programs' verdicts outperforms one trained on a proprietary LLM's labels at two orders of magnitude lower API cost.

cs.AI↗

arXiv2026

Evaluating Sample Utility for Efficient Data Selection by Mimicking Model Weights

Large-scale web-crawled datasets contain noise, bias, and irrelevant information, necessitating data selection techniques. Existing methods depend on hand-crafted heuristics, downstream datasets, or require expensive influence-based computations -- all of which limit scalability and introduce unwanted data dependencies. To address this, we introduce the Mimic Score, a simple and geometry-based data-quality metric that evaluates utility by measuring alignment between a sample's gradients and a target direction induced by a pre-trained reference model. This leverages readily available model weights, avoids needing validation datasets, and incurs minimal computational overheads. Building on this metric, we propose Grad-Mimic, a two-stage framework that re-weights samples online to accelerate training and aggregates sample utilities offline to construct effective data filters. Empirically, we show that using mimic scores to guide training improves data efficiency, accelerates convergence, yields consistent performance gains across six image datasets, and enhances CLIP models with 20.7% fewer training steps. Additionally, mimic score-based filters augment existing filtering techniques, enabling improved CLIP models trained with 4.7 million fewer samples.

cs.LG↗

arXiv2026

SlopCodeBench: Benchmarking How Coding Agents Degrade Over Long-Horizon Iterative Tasks

Software development is iterative, yet agentic coding benchmarks hide design issues through their single-shot setup. Recent iterative benchmarks attempt to remedy this but heavily constrain an agent's design decision space, making it impossible to faithfully measure how their decisions shape future extensions. We introduce SlopCodeBench, a benchmark of 36 problems and 196 checkpoints where agents repeatedly extend their own solutions. Unlike prior iterative benchmarks, our evolving specifications demand architectural decisions but leave internal structure to the agent. We measure two forms of degradation: structural erosion (concentrated complexity) and verbosity (redundant code). Evaluating 15 coding agents across open and closed models, we find that no agent fully solves any problem end-to-end, and the best agent passes 14.8% of checkpoints. Quality degrades across checkpoints, with structural erosion rising in 77% of trajectories and verbosity in 75.5%. Compared to 473 open-source Python repositories, agent code is 2.3x more verbose and 2.0x more eroded, and the human repositories degrade less often and by smaller margins across their git histories. Explicit quality guidance reduces initial verbosity and erosion by up to a third, without affecting degradation rates. SlopCodeBench provides the first measurement of code degradation under iterative extension, revealing that agents pass checkpoints while producing code that erodes and bloats with each turn.

cs.SE↗

arXiv2026

Fine-Tuning Small Reasoning Models for Quantum Field Theory

Despite the growing application of Large Language Models (LLMs) to theoretical physics, there is little academic exploration into how domain-specific physics reasoning ability develops while training these models. To investigate this, we perform the first academic fine-tuning study of small (7B-parameter) reasoning models dedicated specifically to theoretical physics. Because open-source verifiable training data required to train such capabilities is scarce, we developed a robust data generation pipeline that can both create synthetic problems and make existing human-authored problems suitable for model training. Selecting Quantum Field Theory (QFT) as our primary domain, we generated over 2,500 synthetic problems alongside a curated collection of human-adapted problems sourced from arXiv and standard pedagogical resources. We conduct both Reinforcement Learning (RL) and Supervised Fine-Tuning (SFT) experiments, benchmarking performance gains as well as generalization to other physics domains. We perform an extensive analysis of model chains-of-though before and after fine-tuning, to understand how reasoning errors evolve during RL and SFT. Finally, we publicly release our data pipeline, verifiable QFT training data, and $\sim$200M tokens of QFT reasoning traces.

cs.LG↗

arXiv2026

RIFT: A RubrIc Failure Mode Taxonomy and Automated Diagnostics

Rubric-based evaluation is widely used in LLM benchmarks and training pipelines for open-ended, less verifiable tasks. While prior work has demonstrated the effectiveness of rubrics using downstream signals such as reinforcement learning outcomes, there remains no principled way to diagnose how a rubric itself fails from such aggregated or downstream signals alone. To address this gap, we introduce RIFT: RubrIc Failure mode Taxonomy, a taxonomy for systematically characterizing failure modes in rubric composition and design. RIFT consists of eight failure modes organized into three high-level categories: Reliability Failures, Content Validity Failures, and Consequential Validity Failures. RIFT is developed using grounded theory by iteratively annotating rubrics drawn from five diverse data sources spanning general instruction following, code generation, creative writing, and expert-level deep research, until no new failure modes are identified. We evaluate the consistency of the taxonomy by measuring agreement among independent human annotators, observing fair agreement overall (87% pairwise agreement and 0.64 average Cohen's kappa). Finally, to support scalable diagnosis, we propose automated rubric quality metrics and show that they align with human failure-mode annotations, achieving up to 0.925 F1.

cs.AI↗

arXiv2026

Learning from Less: Measuring the Effectiveness of RLVR in Low Data and Compute Regimes

Fine-tuning Large Language Models (LLMs) typically relies on large quantities of high-quality annotated data, or questions with well-defined ground truth answers in the case of Reinforcement Learning with Verifiable Rewards (RLVR). While previous work has explored the benefits to model reasoning capabilities by scaling both data and compute used for RLVR, these results lack applicability in many real-world settings where annotated data and accessible compute may be scarce. In this work, we present a comprehensive empirical study of open-source Small Language Model (SLM) performance after RLVR in low data regimes. Across three novel datasets covering number counting problems, graph reasoning, and spatial reasoning, we characterize how model performance scales with dataset size, diversity, and complexity. We demonstrate that (1) procedural datasets allow for fine-grained evaluation and training dataset development with controllable properties (size, diversity, and complexity), (2) under RLVR, models trained on lower complexity tasks can generalize to higher complexity tasks, and (3) training on mixed complexity datasets is associated with the greatest benefits in low data regimes, providing up to 5x sample efficiency versus training on easy tasks. These findings inspire future work on the development of data scaling laws for RLVR and the use of procedural data generators to further understand effective data development for efficient LLM fine-tuning.

cs.AI↗

arXiv2026

LiveResearchBench: A Live Benchmark for User-Centric Deep Research in the Wild

Deep research -- producing comprehensive, citation-grounded reports by searching and synthesizing information from hundreds of live web sources -- marks an important frontier for agentic systems. To rigorously evaluate this ability, four principles are essential: tasks should be (1) user-centric, reflecting realistic information needs, (2) dynamic, requiring up-to-date information beyond parametric knowledge, (3) unambiguous, ensuring consistent interpretation across users, and (4) multi-faceted and search-intensive, requiring search over numerous web sources and in-depth analysis. Existing benchmarks fall short of these principles, often focusing on narrow domains or posing ambiguous questions that hinder fair comparison. Guided by these principles, we introduce LiveResearchBench, a benchmark of 100 expert-curated tasks spanning daily life, enterprise, and academia, each requiring extensive, dynamic, real-time web search and synthesis. Built with over 1,500 hours of human labor, LiveResearchBench provides a rigorous basis for systematic evaluation. To evaluate citation-grounded long-form reports, we introduce DeepEval, a comprehensive suite covering both content- and report-level quality, including coverage, presentation, citation accuracy and association, consistency and depth of analysis. DeepEval integrates four complementary evaluation protocols, each designed to ensure stable assessment and high agreement with human judgments. Using LiveResearchBench and DeepEval, we conduct a comprehensive evaluation of 17 frontier deep research systems, including single-agent web search, single-agent deep research, and multi-agent systems. Our analysis reveals current strengths, recurring failure modes, and key system components needed to advance reliable, insightful deep research. Our code is available at: https://github.com/SalesforceAIResearch/LiveResearchBench.

cs.AI↗