MuteBench evaluates multimodal fusion robustness to modality missing and within-modality missing on 125000 samples from 9 clinical datasets, finding architecture family predicts tolerance better than parameter count.
hub Canonical reference
Curriculum learning, in: Proceedings of the 26th Annual International Conference on Machine Learning, Associa- tion for Computing Machinery, New York, NY, USA
Canonical reference. 80% of citing Pith papers cite this work as background.
42
Pith papers citing it
4,978
external citations · external index
Background
80% of classified citations
hub tools
citation-role summary
background 8
baseline 1
method 1
citation-polarity summary
representative citing papers
DIPS fine-tunes LLMs to output ordered feasible decision vectors approximating Pareto fronts for constrained bi-objective convex problems, reaching 95-98% normalized hypervolume with 0.16s inference.
Equivariant RL agent synthesizes near-optimal Clifford circuits up to 30 qubits with lower two-qubit gate counts than Qiskit baselines.
Chirality emerges in SMILES translation models through an abrupt encoder-centered reorganization of representations after a long plateau, identified via checkpoint analysis and ablation.
Statistical grammar induction shows the GROWING (bottom-up) maturational account of syntactic category acquisition significantly outperforms the INWARD account across three metrics under identical input and learning conditions.
TCDA introduces TC-DAG to filter cross-thread noise while preserving temporal order and D-RoPE to align semantics across layers and reduce distance dilution, achieving state-of-the-art results on two DiaASQ benchmarks.
A self-evolving framework with proposer-solver-generator roles, Solver Token Entropy, and multi-scale internal evaluation improves unified LMMs on understanding and generation tasks using only self-derived consistency signals.
MAGR-BB matches exhaustive search accuracy on multi-agent Blocksworld while reducing hypothesis evaluations by orders of magnitude via RL scoring inside factorized branch-and-bound.
A model-driven genetic algorithm that mutates RL environments as models generates curriculum families; in a Burning Forest grid-world, curricula reach 100% target success where single-environment training fails.
Dmsh is a new multi-agent RL framework that formulates mesh generation as an MDP and uses three coordinated agents plus curriculum learning to produce globally conforming all-quad meshes without post-processing.
HORIZON is a recoverability-governed checkpointed frontier curriculum for on-policy physical-domain scaling on quadruped locomotion that identifies three regularities: uneven widening, non-monotonic composition, and the necessity of joint on-policy interaction.
Per-Frame Deep Sets enables scaling single-sphere to five-sphere transport on a quadruped by performing permutation-invariant pooling within each history frame, reaching 100% no-drop success in simulation where standard encoders plateau.
CDS-trained BabyLMs show earlier and more appropriate production in a new frame-completion task while FineWeb-edu models lead on comprehension benchmarks, indicating current tests underestimate CDS benefits.
A VAE-based latent task representation enables automatic curriculum generation in CRL for non-Euclidean navigation tasks, outperforming interpolation and GAN-based methods in experiments.
Learning-Zone Energy is a new online data selection framework for RL post-training that retains 40% of data per step yet matches or exceeds full-data baselines on math tasks with 36% lower FLOPs.
An auxiliary modulus during training reduces wrap-around issues and preserves train-test input distributions, enabling better accuracy and sample efficiency for large N and q in modular addition learning.
Skill-SD turns an agent's completed trajectories into dynamic natural-language skills that condition only the teacher in self-distillation, yielding 14-42% gains over RL and OPSD baselines on multi-turn agent benchmarks.
A modular framework combining physics-informed neural networks, Ornstein-Uhlenbeck fitting, extreme value theory, and vision-language models detects 51 transient flares in 9,258 SDSS Stripe 82 quasar light curves.
IDCL adds density-based curriculum learning and density-core guidance to deep image clustering, claiming superior robustness, faster convergence, and flexibility on benchmark datasets.
An LLM acting as real-time controller for SIMP topology optimization parameters outperforms fixed schedules and heuristics, delivering 5.7-18.1% lower compliance on 2D and 3D benchmarks.
CausalTimePrior generates synthetic temporal structural causal models with paired observational and interventional time series to train prior-data fitted networks for in-context causal effect estimation on held-out data.
RAG-GNN augments GNNs with retrieved literature knowledge via gated fusion to improve functional clustering of 379 proteins in cancer signaling networks, raising silhouette score by 0.093.
SAM 3D reconstructs 3D objects from single images with geometry, texture, and pose using human-model annotated data at scale and synthetic-to-real training, achieving 5:1 human preference wins.
CL-MARL uses an adaptive curriculum scheduler called FlexDiff and Counterfactual Group Relative Policy Advantage to break static-difficulty training in MARL and achieve higher win rates on hard StarCraft maps.
citing papers explorer
-
MuteBench: Modality Unavailability Tolerance Evaluation for Incomplete Multimodal Fusion
MuteBench evaluates multimodal fusion robustness to modality missing and within-modality missing on 125000 samples from 9 clinical datasets, finding architecture family predicts tolerance better than parameter count.
-
Large Language Models as Amortized Pareto-Front Generators for Constrained Bi-Objective Convex Optimization
DIPS fine-tunes LLMs to output ordered feasible decision vectors approximating Pareto fronts for constrained bi-objective convex problems, reaching 95-98% normalized hypervolume with 0.16s inference.
-
Equivariant Reinforcement Learning for Clifford Quantum Circuit Synthesis
Equivariant RL agent synthesizes near-optimal Clifford circuits up to 30 qubits with lower two-qubit gate counts than Qiskit baselines.
-
From Syntax to Semantics: Unveiling the Emergence of Chirality in SMILES Translation Models
Chirality emerges in SMILES translation models through an abrupt encoder-centered reorganization of representations after a long plateau, identified via checkpoint analysis and ablation.
-
A Computational Operationalisation of Competing Maturational Theories of Syntactic Development via Statistical Grammar Induction
Statistical grammar induction shows the GROWING (bottom-up) maturational account of syntactic category acquisition significantly outperforms the INWARD account across three metrics under identical input and learning conditions.
-
TCDA: Thread-Constrained Discourse-Aware Modeling for Conversational Sentiment Quadruple Analysis
TCDA introduces TC-DAG to filter cross-thread noise while preserving temporal order and D-RoPE to align semantics across layers and reduce distance dilution, achieving state-of-the-art results on two DiaASQ benchmarks.
-
Ask, Solve, Generate: Self-Evolving Unified Multimodal Understanding and Generation via Self-Consistency Rewards
A self-evolving framework with proposer-solver-generator roles, Solver Token Entropy, and multi-scale internal evaluation improves unified LMMs on understanding and generation tasks using only self-derived consistency signals.
-
Multi-Agent Goal Recognition with Team- and Goal-Conditioned Reinforcement Learning and Factorized Branch-and-Bound
MAGR-BB matches exhaustive search accuracy on multi-agent Blocksworld while reducing hypothesis evaluations by orders of magnitude via RL scoring inside factorized branch-and-bound.
-
A Model-Driven Approach for Developing Families of Reinforcement Learning Environments
A model-driven genetic algorithm that mutates RL environments as models generates curriculum families; in a Burning Forest grid-world, curricula reach 100% target success where single-environment training fails.
-
Dmsh: A Multi-Agent Reinforcement Learning Framework for All-Quad Mesh Generation
Dmsh is a new multi-agent RL framework that formulates mesh generation as an MDP and uses three coordinated agents plus curriculum learning to produce globally conforming all-quad meshes without post-processing.
-
HORIZON: Recoverability-Governed Curriculum for Physical-Domain Scaling
HORIZON is a recoverability-governed checkpointed frontier curriculum for on-policy physical-domain scaling on quadruped locomotion that identifies three regularities: uneven widening, non-monotonic composition, and the necessity of joint on-policy interaction.
-
S2M-Trek: From Single to Multi-Sphere Transport via Per-Frame Deep Sets on a Wheel-Legged Robot
Per-Frame Deep Sets enables scaling single-sphere to five-sphere transport on a quadruped by performing permutation-invariant pooling within each history frame, reaching 100% no-drop success in simulation where standard encoders plateau.
-
Child-directed speech facilitates production, not comprehension, in BabyLMs
CDS-trained BabyLMs show earlier and more appropriate production in a new frame-completion task while FineWeb-edu models lead on comprehension benchmarks, indicating current tests underestimate CDS benefits.
-
Curriculum reinforcement learning with measurable task representation learning
A VAE-based latent task representation enables automatic curriculum generation in CRL for non-Euclidean navigation tasks, outperforming interpolation and GAN-based methods in experiments.
-
Learning-Zone Energy: Online Data Selection for Efficient RL Post-Training
Learning-Zone Energy is a new online data selection framework for RL post-training that retains 40% of data per step yet matches or exceeds full-data baselines on math tasks with 36% lower FLOPs.
-
Learning Large-Scale Modular Addition with an Auxiliary Modulus
An auxiliary modulus during training reduces wrap-around issues and preserves train-test input distributions, enabling better accuracy and sample efficiency for large N and q in modular addition learning.
-
Skill-SD: Skill-Conditioned Self-Distillation for Multi-turn LLM Agents
Skill-SD turns an agent's completed trajectories into dynamic natural-language skills that condition only the teacher in self-distillation, yielding 14-42% gains over RL and OPSD baselines on multi-turn agent benchmarks.
-
A Statistical-AI Framework for Detecting Transient Flares in SDSS Stripe 82 Quasar Light Curves
A modular framework combining physics-informed neural networks, Ornstein-Uhlenbeck fitting, extreme value theory, and vision-language models detects 51 transient flares in 9,258 SDSS Stripe 82 quasar light curves.
-
Deep Image Clustering Based on Curriculum Learning and Density Information
IDCL adds density-based curriculum learning and density-core guidance to deep image clustering, claiming superior robustness, faster convergence, and flexibility on benchmark datasets.
-
Large Language Models as Optimization Controllers: Adaptive Continuation for SIMP Topology Optimization
An LLM acting as real-time controller for SIMP topology optimization parameters outperforms fixed schedules and heuristics, delivering 5.7-18.1% lower compliance on 2D and 3D benchmarks.
-
Interventional Time Series Priors for Causal Foundation Models
CausalTimePrior generates synthetic temporal structural causal models with paired observational and interventional time series to train prior-data fitted networks for in-context causal effect estimation on held-out data.
-
RAG-GNN: Integrating Retrieved Knowledge with Graph Neural Networks for Precision Medicine
RAG-GNN augments GNNs with retrieved literature knowledge via gated fusion to improve functional clustering of 379 proteins in cancer signaling networks, raising silhouette score by 0.093.
-
SAM 3D: 3Dfy Anything in Images
SAM 3D reconstructs 3D objects from single images with geometry, texture, and pose using human-model annotated data at scale and synthetic-to-real training, achieving 5:1 human preference wins.
-
Overcoming Environmental Meta-Stationarity in MARL via Adaptive Curriculum and Counterfactual Group Advantage
CL-MARL uses an adaptive curriculum scheduler called FlexDiff and Counterfactual Group Relative Policy Advantage to break static-difficulty training in MARL and achieve higher win rates on hard StarCraft maps.
-
Remote Rowhammer Attack using Adversarial Observations on Federated Learning Clients
A reinforcement learning attacker manipulates client sensor observations in federated learning to induce repetitive server memory updates, achieving around 70% repeated update rate and enabling remote Rowhammer bit flips on an automatic speech recognition model.
-
CODA: A Continuous Online Evolve Framework for Deploying HAR Sensing Systems
CODA enables continuous online adaptation for HAR sensing by cache-based selective assimilation of informative instances and adaptive temporal retention to forget obsolete data under non-stationary drift.
-
UASPL: Uncertainty-Aware Self-Paced Learning with Evidential Neural Networks
UASPL selects training samples by an evidential loss that couples label fit with adaptive uncertainty-weighted KL regularization, yielding an easy-to-hard preference and stronger average classification results than loss-only SPL.
-
Fair Cognitive Impairment Detection Through Unlearning
A multimodal fusion plus gradient-reversal unlearning framework improves MCI classification accuracy and reduces performance gaps across sex and language subgroups on TAUKADIAL and PREPARE.
-
ReCal: Reward Calibration for RL-based LLM Routing
ReCal introduces hierarchical reward decomposition and distribution-aware optimization to address ambiguous credit assignment and optimization bias in RL-based LLM routing.
-
MoEIoU: Rethinking Bounding-Box Regression as a Mixture of Experts
MoEIoU is a mixture-of-experts IoU loss using log-sum-exp aggregation and curriculum weighting that reports consistent gains over prior IoU losses on PASCAL VOC, HRIPCB, and MS COCO with YOLO models.
-
Automating Formal Verification with Reinforcement Learning and Recursive Inference
RLVR training raises verified Dafny pass rates from 9.7% to 31.1% on a filtered benchmark while a Lean proof scaffold lifts success from 46.2% to 69.2% on a pilot set and solves 7 of 42 prior unsolved tasks.
-
Closed-Loop Sim-to-Real Reinforcement Learning for Deformable Microfiber Shape Control
A closed-loop sim-to-real RL policy trained in a simplified frictionless simulator achieves sub-millimeter microfiber shape control on physical hardware via visual feedback without retraining.
-
Consistently Informative Soft-Label Temperature for Knowledge Distillation
CIST uses per-sample adaptive temperatures for both teacher and student in knowledge distillation to ensure consistent entropy in soft labels and reports gains on vision and language tasks.
-
Mistake gating leads to energy and memory efficient continual learning
Memorized mistake-gated learning updates synapses only on current or past classification errors, cutting parameter updates by 50–80% with little accuracy loss and smaller replay buffers.
-
A Comprehensive Survey of Agents for Computer Use: Foundations, Challenges, and Future Directions
A survey of 87 agents for computer use and 33 datasets that introduces a three-dimensional taxonomy across domain, interaction, and agent perspectives and identifies six research gaps.
-
Data-CUBE: Data Curriculum for Instruction-based Sentence Representation Learning
Data-CUBE applies a two-level curriculum (TSP-based task ordering via simulated annealing plus difficulty-sorted mini-batches) to multi-task instruction tuning and reports gains on MTEB sentence representation tasks.
-
SkillChain: Closing the Loop on Skill Evolution for Image-Based E-Commerce AI Assistants
SkillChain automates skill lifecycle for e-commerce image AI assistants via creator, optimizer, and refiner stages, leading to improved response quality and user engagement in production A/B tests.
-
Compatibility-Aware Dynamic Fine-Tuning for Large Language Models
CADFT improves supervised fine-tuning of large language models by dynamically down-weighting training samples whose low model-likelihood indicates high gradient variance, yielding better stability and generalization.
-
AI-Based Detection of Temporal Changes in MR-Linac Images Acquired During Routine Prostate Radiotherapy
Deep learning on pairwise comparisons detects temporal changes in 0.35T MR-Linac images during prostate radiotherapy with AUC 0.99 for first-to-last fraction pairs.
-
Robust Training for Speaker Verification against Noisy Labels
A two-stage iterative method with OR-Gate top-k selection filters noisy labels during training of speaker verification models on VoxCeleb datasets.
-
How Complexity Contributes to Learning Opacity in Machine Learning
Neural network learning opacity stems from three dynamical complexity properties in training, rendering some sources of opacity irreducible.
-
Zero-shot Transfer of Reinforcement Learning Control Policies for the Swing-Up and Stabilization of a Cart-Pole System
Zero-shot sim-to-real transfer of independently trained RL policies for cart-pole swing-up and stabilization is achieved via sensitivity-guided domain randomization, linear curriculum learning, and first-order action smoothing with Simulink switching logic.