Mila

LLMs possess an internal estimate of their remaining output length, revealing a surprising layer of planning in their generation process.

Self-distillation may boost accuracy but comes at the hidden cost of significantly reduced output diversity, risking performance in diverse scenarios.

Relying on ChatGPT for information seeking may diminish users' agency and critical learning outcomes, revealing hidden risks of generative AI in education.

STORM transforms lexical query expansion by turning retrieval rewards into actionable token-level signals, enabling efficient and effective query rewriting that rivals larger models.

Bridging the scientific knowledge gap for hundreds of millions, AfriScience-MT pioneers document-level scientific machine translation for six African languages.

Language specialization in multilingual MoEs happens mostly in the final layers, suggesting a surprisingly simple recipe for parameter-efficient adaptation.

Human-generated citation lists, long considered the gold standard for evaluating literature search, are surprisingly unreliable, with LLMs judging them relevant only ~50% of the time.

LLMs struggle with structured 2D tasks when inputs are serialized into 1D, revealing a surprising performance gap compared to vision-augmented models that directly process the 2D layout.

CroSearch-R1 reveals that integrating cross-lingual knowledge through a dynamic retrieval strategy can substantially enhance the performance of Retrieval-Augmented Generation systems.

ManifoldRank reveals that treating fairness as a taxation cost can significantly enhance the effectiveness of online fair re-ranking algorithms.

LLMs re-rank documents better when you learn to route each query to the specific attention heads that matter, instead of relying on static subsets or everything at once.

Continuously nudging LLM activations during generation can effectively correct misalignment without sacrificing coherence, offering a lightweight runtime defense against adversarial prompts and other triggers.

Ditch the cross-world counterfactuals: Sequential Transport offers a DAG-aware, optimal transport approach to causal mediation analysis, providing deterministic counterfactual mediators and fine-grained attribution.

One in four initial posts on a major cybercrime forum contain explicit crime-related content, revealing a surprisingly high baseline of open criminal activity.

Diagonal SSMs, despite their empirical success, provably fail to track states of non-Abelian groups, revealing fundamental limitations in their expressive power.

Forget full fine-tuning: this dynamic routing strategy lets you adapt dense retrieval to new domains while using just 2% of the parameters.

Achieve state-of-the-art dynamic graph anomaly detection with limited labels by learning a robust decision boundary around normal data, outperforming methods that overfit to scarce anomalies.

Attention-based re-ranking gets a boost: ReAttn's post-hoc re-weighting tames over-concentration and lexical bias, leading to more accurate and interpretable results without extra training.

LLMs struggle to balance rational financial decisions with mimicking noisy user behavior, often overfitting to short-term market trends instead of aligning with long-term investment goals.

Cybercriminals are actively exploring AI's potential for both enhancing existing attacks and creating novel illicit tools, but harbor significant doubts about its real-world effectiveness and impact on their operations.

Dramatically improve protein language models by simply post-training them to align with protein graphs, yielding a 59% increase in contact prediction accuracy.