Boosting LLM Reasoning via Spontaneous Self-Correction

While large language models (LLMs) have demonstrated remarkable success on a broad range of tasks, math reasoning remains a challenging one. One of the approaches for improving math reasoning is self-correction, which designs self-improving loops to let the model correct its own mistakes. However, existing self-correction approaches treat corrections as standalone post-generation refinements, relying on extra prompt and system designs to elicit self-corrections, instead of performing real-time, spontaneous self-corrections in a single pass. To address this, we propose SPOC, a spontaneous self-correction approach that enables LLMs to generate interleaved solutions and verifications in a single inference pass, with generation dynamically terminated based on verification outcomes, thereby effectively scaling inference time compute. SPOC considers a multi-agent perspective by assigning dual roles -- solution proposer and verifier -- to the same model. We adopt a simple yet effective approach to generate synthetic data for fine-tuning, enabling the model to develop capabilities for self-verification and multi-agent collaboration. We further improve its solution proposal and verification accuracy through online reinforcement learning. Experiments on mathematical reasoning benchmarks show that SPOC significantly improves performance. Notably, SPOC boosts the accuracy of Llama-3.1-8B and 70B Instruct models, achieving gains of 8.8% and 11.6% on MATH500, 10.0% and 20.0% on AMC23, and 3.3% and 6.7% on AIME24, respectively.

Reinforcement Learning from User Feedback

As large language models (LLMs) are increasingly deployed in diverse user facing applications, aligning them with real user preferences becomes essential. Existing methods like Reinforcement Learning from Human Feedback (RLHF) rely on expert annotators trained on manually defined guidelines, whose judgments may not reflect the priorities of everyday users. We introduce Reinforcement Learning from User Feedback (RLUF), a framework for aligning LLMs directly to implicit signals from users in production. RLUF addresses key challenges of user feedback: user feedback is often binary (e.g., emoji reactions), sparse, and occasionally adversarial. We train a reward model, P[Love], to predict the likelihood that an LLM response will receive a Love Reaction, a lightweight form of positive user feedback, and integrate P[Love] into a multi-objective policy optimization framework alongside helpfulness and safety objectives. In large-scale experiments, we show that P[Love] is predictive of increased positive feedback and serves as a reliable offline evaluator of future user behavior. Policy optimization using P[Love] significantly raises observed positive-feedback rates, including a 28% increase in Love Reactions during live A/B tests. However, optimizing for positive reactions introduces reward hacking challenges, requiring careful balancing of objectives. By directly leveraging implicit signals from users, RLUF offers a path to aligning LLMs with real-world user preferences at scale.

Learning Auxiliary Tasks Improves Reference-Free Hallucination Detection in Open-Domain Long-Form Generation

Hallucination, the generation of factually incorrect information, remains a significant challenge for large language models (LLMs), especially in open-domain long-form generation. Existing approaches for detecting hallucination in long-form tasks either focus on limited domains or rely heavily on external fact-checking tools, which may not always be available. In this work, we systematically investigate reference-free hallucination detection in open-domain long-form responses. Our findings reveal that internal states (e.g., model's output probability and entropy) alone are insufficient for reliably (i.e., better than random guessing) distinguishing between factual and hallucinated content. To enhance detection, we explore various existing approaches, including prompting-based methods, probing, and fine-tuning, with fine-tuning proving the most effective. To further improve the accuracy, we introduce a new paradigm, named RATE-FT, that augments fine-tuning with an auxiliary task for the model to jointly learn with the main task of hallucination detection. With extensive experiments and analysis using a variety of model families & datasets, we demonstrate the effectiveness and generalizability of our method, e.g., +3% over general fine-tuning methods on LongFact.

Sim-to-Real: An Unsupervised Noise Layer for Screen-Camera Watermarking Robustness

Unauthorized screen capturing and dissemination pose severe security threats such as data leakage and information theft. Several studies propose robust watermarking methods to track the copyright of Screen-Camera (SC) images, facilitating post-hoc certification against infringement. These techniques typically employ heuristic mathematical modeling or supervised neural network fitting as the noise layer, to enhance watermarking robustness against SC. However, both strategies cannot fundamentally achieve an effective approximation of SC noise. Mathematical simulation suffers from biased approximations due to the incomplete decomposition of the noise and the absence of interdependence among the noise components. Supervised networks require paired data to train the noise-fitting model, and it is difficult for the model to learn all the features of the noise. To address the above issues, we propose Simulation-to-Real (S2R). Specifically, an unsupervised noise layer employs unpaired data to learn the discrepancy between the modeling simulated noise distribution and the real-world SC noise distribution, rather than directly learning the mapping from sharp images to real-world images. Learning this transformation from simulation to reality is inherently simpler, as it primarily involves bridging the gap in noise distributions, instead of the complex task of reconstructing fine-grained image details. Extensive experimental results validate the efficacy of the proposed method, demonstrating superior watermark robustness and generalization compared to those of state-of-the-art methods.

Gaussian Shading++: Rethinking the Realistic Deployment Challenge of Performance-Lossless Image Watermark for Diffusion Models

Ethical concerns surrounding copyright protection and inappropriate content generation pose challenges for the practical implementation of diffusion models. One effective solution involves watermarking the generated images. Existing methods primarily focus on ensuring that watermark embedding does not degrade the model performance. However, they often overlook critical challenges in real-world deployment scenarios, such as the complexity of watermark key management, user-defined generation parameters, and the difficulty of verification by arbitrary third parties. To address this issue, we propose Gaussian Shading++, a diffusion model watermarking method tailored for real-world deployment. We propose a double-channel design that leverages pseudorandom error-correcting codes to encode the random seed required for watermark pseudorandomization, achieving performance-lossless watermarking under a fixed watermark key and overcoming key management challenges. Additionally, we model the distortions introduced during generation and inversion as an additive white Gaussian noise channel and employ a novel soft decision decoding strategy during extraction, ensuring strong robustness even when generation parameters vary. To enable third-party verification, we incorporate public key signatures, which provide a certain level of resistance against forgery attacks even when model inversion capabilities are fully disclosed. Extensive experiments demonstrate that Gaussian Shading++ not only maintains performance losslessness but also outperforms existing methods in terms of robustness, making it a more practical solution for real-world deployment.

Tracking-Aware Deformation Field Estimation for Non-rigid 3D Reconstruction in Robotic Surgeries

Minimally invasive procedures have been advanced rapidly by the robotic laparoscopic surgery. The latter greatly assists surgeons in sophisticated and precise operations with reduced invasiveness. Nevertheless, it is still safety critical to be aware of even the least tissue deformation during instrument-tissue interactions, especially in 3D space. To address this, recent works rely on NeRF to render 2D videos from different perspectives and eliminate occlusions. However, most of the methods fail to predict the accurate 3D shapes and associated deformation estimates robustly. Differently, we propose Tracking-Aware Deformation Field (TADF), a novel framework which reconstructs the 3D mesh along with the 3D tissue deformation simultaneously. It first tracks the key points of soft tissue by a foundation vision model, providing an accurate 2D deformation field. Then, the 2D deformation field is smoothly incorporated with a neural implicit reconstruction network to obtain tissue deformation in the 3D space. Finally, we experimentally demonstrate that the proposed method provides more accurate deformation estimation compared with other 3D neural reconstruction methods in two public datasets.

Think Smarter not Harder: Adaptive Reasoning with Inference Aware Optimization

Solving mathematics problems has been an intriguing capability of large language models, and many efforts have been made to improve reasoning by extending reasoning length, such as through self-correction and extensive long chain-of-thoughts. While promising in problem-solving, advanced long reasoning chain models exhibit an undesired single-modal behavior, where trivial questions require unnecessarily tedious long chains of thought. In this work, we propose a way to allow models to be aware of inference budgets by formulating it as utility maximization with respect to an inference budget constraint, hence naming our algorithm Inference Budget-Constrained Policy Optimization (IBPO). In a nutshell, models fine-tuned through IBPO learn to ``understand'' the difficulty of queries and allocate inference budgets to harder ones. With different inference budgets, our best models are able to have a $4.14$\% and $5.74$\% absolute improvement ($8.08$\% and $11.2$\% relative improvement) on MATH500 using $2.16$x and $4.32$x inference budgets respectively, relative to LLaMA3.1 8B Instruct. These improvements are approximately $2$x those of self-consistency under the same budgets.

ASAP: Learning Generalizable Online Bin Packing via Adaptive Selection After Pruning

Recently, deep reinforcement learning (DRL) has achieved promising results in solving online 3D Bin Packing Problems (3D-BPP). However, these DRL-based policies may perform poorly on new instances due to distribution shift. Besides generalization, we also consider adaptation, completely overlooked by previous work, which aims at rapidly finetuning these policies to a new test distribution. To tackle both generalization and adaptation issues, we propose Adaptive Selection After Pruning (ASAP), which decomposes a solver's decision-making into two policies, one for pruning and one for selection. The role of the pruning policy is to remove inherently bad actions, which allows the selection policy to choose among the remaining most valuable actions. To learn these policies, we propose a training scheme based on a meta-learning phase of both policies followed by a finetuning phase of the sole selection policy to rapidly adapt it to a test distribution. Our experiments demonstrate that ASAP exhibits excellent generalization and adaptation capabilities on in-distribution and out-of-distribution instances under both discrete and continuous setup.

Step-KTO: Optimizing Mathematical Reasoning through Stepwise Binary Feedback

Large language models (LLMs) have recently demonstrated remarkable success in mathematical reasoning. Despite progress in methods like chain-of-thought prompting and self-consistency sampling, these advances often focus on final correctness without ensuring that the underlying reasoning process is coherent and reliable. This paper introduces Step-KTO, a training framework that combines process-level and outcome-level binary feedback to guide LLMs toward more trustworthy reasoning trajectories. By providing binary evaluations for both the intermediate reasoning steps and the final answer, Step-KTO encourages the model to adhere to logical progressions rather than relying on superficial shortcuts. Our experiments on challenging mathematical benchmarks show that Step-KTO significantly improves both final answer accuracy and the quality of intermediate reasoning steps. For example, on the MATH-500 dataset, Step-KTO achieves a notable improvement in Pass@1 accuracy over strong baselines. These results highlight the promise of integrating stepwise process feedback into LLM training, paving the way toward more interpretable and dependable reasoning capabilities.

END$^2$: Robust Dual-Decoder Watermarking Framework Against Non-Differentiable Distortions

DNN-based watermarking methods have rapidly advanced, with the ``Encoder-Noise Layer-Decoder'' (END) framework being the most widely used. To ensure end-to-end training, the noise layer in the framework must be differentiable. However, real-world distortions are often non-differentiable, leading to challenges in end-to-end training. Existing solutions only treat the distortion perturbation as additive noise, which does not fully integrate the effect of distortion in training. To better incorporate non-differentiable distortions into training, we propose a novel dual-decoder architecture (END$^2$). Unlike conventional END architecture, our method employs two structurally identical decoders: the Teacher Decoder, processing pure watermarked images, and the Student Decoder, handling distortion-perturbed images. The gradient is backpropagated only through the Teacher Decoder branch to optimize the encoder thus bypassing the problem of non-differentiability. To ensure resistance to arbitrary distortions, we enforce alignment of the two decoders' feature representations by maximizing the cosine similarity between their intermediate vectors on a hypersphere. Extensive experiments demonstrate that our scheme outperforms state-of-the-art algorithms under various non-differentiable distortions. Moreover, even without the differentiability constraint, our method surpasses baselines with a differentiable noise layer. Our approach is effective and easily implementable across all END architectures, enhancing practicality and generalizability.