Author

Benjamin Van Durme

h-index 18 View on Semantic Scholar

We have 6 of 83 papers

We haven’t gathered this author’s papers yet. Follow them and we’ll fetch their work.

Not the right person? Other researchers publish under this name.

Preprint Oct 2026

FOCUS: Active Critic Learning for Long-Horizon Language Model Agents

Sparse outcome rewards make it difficult for long-horizon language-model agents to assign credit to earlier actions. Learned critics provide value estimates for intermediate states, but these estimates can be inaccurate, while estimating intermediate values with repetitive continuation samples from points within a traj...

Zhengping Jiang, William Jurayj, Liaoyaqi Wang et al. · 0 citations

#software testing Preprint Oct 2026

SIGMA: Self-Improving Alignment Generalization from a Model Spec

SIGMA is proposed, a data generation and training pipeline enabling alignment self-improvement that generalizes to out-of-distribution settings and shows that a Model Spec balancing harmlessness and helpfulness, test-time reasoning for safety deliberation, and high-quality rubrics from SIGMA's task designer agent are c...

Jing-Yu Zhang, Shruti Palaskar, Daniel Khashabi et al. · 0 citations

#machine learning Apr 2026

Unified Multimodal Uncertain Inference

CLUE (Calibrated Latent Uncertainty Estimation), which combines self-consistent teacher calibration and distribution-based confidence probing to produce calibrated predictions, is introduced, which achieves equivalent or stronger performance than zero-shot baselines up to 32B parameters across all modalities.

Dengjia Zhang, Alexander Martin, William Jurayj et al. · arXiv.org · 4 citations

#artificial intelligence Apr 2026

Many-Tier Instruction Hierarchy in LLM Agents

This work proposes Many-Tier Instruction Hierarchy (ManyIH), a paradigm for resolving instruction conflicts among instructions with arbitrarily many privilege levels, and introduces ManyIH-Bench, the first benchmark for ManyIH.

Jing-Yu Zhang, Tian-Jian Li, William Jurayj et al. · arXiv.org · 9 citations

#artificial intelligence May 2026

Can Coding Agents Reproduce Findings in Computational Materials Science?

This work presents AutoMat, a benchmark for evaluating LLM-based agents'ability to reproduce claims from computational materials science, and curates a set of claims from real materials science papers to test whether coding agents can recover and execute the end-to-end workflow needed to support (or undermine) such cla...

Ziyang Huang, Yiyun Cao, A. Shargh et al. · arXiv.org · 9 citations

Preprint Aug 2026

Information Abundance Paradox: Long-Context Training Undermines Parametric Knowledge

The Information Abundance Paradox is supported and it is suggested that scaling toward near-infinite context is not simply a matter of supplying more data, even when high-quality long-context data is abundant.

Arda Uzunouglu, Benjamin Van Durme, Daniel Khashabi · 2 citations