PARAMΔ upcycles dense models to MoE for per-language experts and grafts post-training deltas to enable data-efficient language expansion while preserving original capabilities.
hub
Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations) , address=
15 Pith papers cite this work. Polarity classification is still indexing.
15
Pith papers citing it
hub tools
citation-role summary
background 1
citation-polarity summary
years
2026 15roles
background 1polarities
background 1representative citing papers
StepCodeReasoner aligns code reasoning with verifiable stepwise execution traces via print anchors and bi-level GRPO reinforcement learning, reaching SOTA results on CRUXEval (91.1%) and LiveCodeBench (86.5%) for a 7B model.
UniCoder applies symbolic attribute alignment via an auxiliary LLM and reference-guided optimization in RL to achieve SOTA visual-to-code generation on ChartMimic, UniSVG, Design2Code, and ScreenBench.
A dual hierarchical RL framework with two agents coordinates high-level dialogue strategy and low-level question generation to emulate judicial questioning and extract key information from Supreme Court arguments, outperforming baselines.
Imagining in 360° decouples visual search into a single-step probabilistic semantic layout predictor and an actor, removing the need for multi-turn CoT reasoning and trajectory annotations while improving efficiency in 360° environments.
DoTS decouples SFT and RLVR training then synthesizes their task vectors at inference time to match integrated training results at ~3% compute cost.
SafetyALFRED shows multimodal LLMs recognize kitchen hazards accurately in QA tests but achieve low success rates when required to mitigate those hazards through embodied planning.
CausalMix fits a causal model on 512 runs of a 0.5B model to estimate CATE, then extrapolates optimal mixtures for an 800K data pool applied to 7B and 4B models, outperforming RegMix.
TIF-GRPO uses integral feedback on pseudo-temporal trajectories to regulate anatomy-aware rewards in RL for clinical faithfulness in volumetric CT analysis.
On-policy distillation gains efficiency from early foresight in module allocation and update directions, which the proposed EffOPD method exploits for 3x faster training with comparable performance.
UserGPT introduces a generative LLM framework with a behavior simulation engine, semantization module, and DF-GRPO post-training that scores 0.7325 on tag prediction and 0.7528 on summary generation on HPR-Bench while compressing records by up to 97.9%.
SimpleSearch-VL improves Qwen3-VL multimodal agent baselines by 15.8-16 points on average using 7K total training examples and reaches parity with Gemini-3-Pro on the 30B variant.
Proposes Modality-Aware Credit Assignment (MoCA) with blindfolded-reasoning proxy to reward perception fidelity separately from reasoning in VLMs.
citing papers explorer
-
A Data-Efficient Path to Multilingual LLMs: Language Expansion via Post-training PARAM$\Delta$ Integration into Upcycled MoE
PARAMΔ upcycles dense models to MoE for per-language experts and grafts post-training deltas to enable data-efficient language expansion while preserving original capabilities.
-
StepCodeReasoner: Aligning Code Reasoning with Stepwise Execution Traces via Reinforcement Learning
StepCodeReasoner aligns code reasoning with verifiable stepwise execution traces via print anchors and bi-level GRPO reinforcement learning, reaching SOTA results on CRUXEval (91.1%) and LiveCodeBench (86.5%) for a 7B model.
-
UniCoder: Unified Visual-to-Code Generation via Symbolic Rewards and Reference-Guided Code Optimization
UniCoder applies symbolic attribute alignment via an auxiliary LLM and reference-guided optimization in RL to achieve SOTA visual-to-code generation on ChartMimic, UniSVG, Design2Code, and ScreenBench.
-
Dual Hierarchical Dialogue Policy Learning for Legal Inquisitive Conversational Agents
A dual hierarchical RL framework with two agents coordinates high-level dialogue strategy and low-level question generation to emulate judicial questioning and extract key information from Supreme Court arguments, outperforming baselines.
-
Beyond Thinking: Imagining in 360$^\circ$ for Humanoid Visual Search
Imagining in 360° decouples visual search into a single-step probabilistic semantic layout predictor and an actor, removing the need for multi-turn CoT reasoning and trajectory annotations while improving efficiency in 360° environments.
-
Decouple before Integration: Test-time Synthesis of SFT and RLVR Task Vectors
DoTS decouples SFT and RLVR training then synthesizes their task vectors at inference time to match integrated training results at ~3% compute cost.
-
SafetyALFRED: Evaluating Safety-Conscious Planning of Multimodal Large Language Models
SafetyALFRED shows multimodal LLMs recognize kitchen hazards accurately in QA tests but achieve low success rates when required to mitigate those hazards through embodied planning.
-
CausalMix: Data Mixture as Causal Inference for Language Model Training
CausalMix fits a causal model on 512 runs of a 0.5B model to estimate CATE, then extrapolates optimal mixtures for an 800K data pool applied to 7B and 4B models, outperforming RegMix.
-
Regulating Anatomy-Aware Rewards via Trajectory-Integral Feedback for Volumetric Computed Tomography Analysis
TIF-GRPO uses integral feedback on pseudo-temporal trajectories to regulate anatomy-aware rewards in RL for clinical faithfulness in volumetric CT analysis.
-
Learning to Foresee: Unveiling the Unlocking Efficiency of On-Policy Distillation
On-policy distillation gains efficiency from early foresight in module allocation and update directions, which the proposed EffOPD method exploits for 3x faster training with comparable performance.
-
UserGPT Technical Report
UserGPT introduces a generative LLM framework with a behavior simulation engine, semantization module, and DF-GRPO post-training that scores 0.7325 on tag prediction and 0.7528 on summary generation on HPR-Bench while compressing records by up to 97.9%.
-
SimpleSearch-VL: A Simple Recipe for Multimodal Agentic Deep Search
SimpleSearch-VL improves Qwen3-VL multimodal agent baselines by 15.8-16 points on average using 7K total training examples and reaches parity with Gemini-3-Pro on the 30B variant.
-
Bad Seeing or Bad Thinking? Rewarding Perception for Multimodal Reasoning
Proposes Modality-Aware Credit Assignment (MoCA) with blindfolded-reasoning proxy to reward perception fidelity separately from reasoning in VLMs.
- Message Passing Enables Efficient Reasoning
- Hybrid Policy Distillation for LLMs