On Two Open Problems by Johnson and Koll\'ar
We resolve two problems left open by Johnson and Koll\'ar in their paper Arcology: Question 21 and the openness issue arising in Example 14.
math.AG↗
arXiv subjects
Siyuan Fu
Publications and source records attributed to Siyuan Fu.
FP4 Explore, BF16 Train: Diffusion Reinforcement Learning via Efficient Rollout Scaling
Reinforcement-Learning-based post-training has recently emerged as a promising paradigm for aligning text-to-image diffusion models with human preferences. In recent studies, increasing the rollout group size yields pronounced performance improvements, indicating substantial room for further alignment gains. However, scaling rollouts on large-scale foundational diffusion models (e.g., FLUX.1-12B) imposes a heavy computational burden. To alleviate this bottleneck, we explore the integration of FP4 quantization into Diffusion RL rollouts. Yet, we identify that naive quantized pipelines inherently introduce risks of performance degradation. To overcome this dilemma between efficiency and training integrity, we propose Sol-RL (Speed-of-light RL), a novel FP4-empowered Two-stage Reinforcement Learning framework. First, we utilize high-throughput NVFP4 rollouts to generate a massive candidate pool and extract a highly contrastive subset. Second, we regenerate these selected samples in BF16 precision and optimize the policy exclusively on them. By decoupling candidate exploration from policy optimization, Sol-RL integrates the algorithmic mechanisms of rollout scaling with the system-level throughput gains of NVFP4. This synergistic algorithm-hardware design effectively accelerates the rollout phase while reserving high-fidelity samples for optimization. We empirically demonstrate that our framework maintains the training integrity of BF16 precision pipeline while fully exploiting the throughput gains enabled by FP4 arithmetic. Extensive experiments across SANA, FLUX.1, and SD3.5-L substantiate that our approach delivers superior alignment performance across multiple metrics while accelerating training convergence by up to $4.64\times$, unlocking the power of massive rollout scaling at a fraction of the cost.
cs.LG↗
SafeSeek: Universal Attribution of Safety Circuits in Language Models
Mechanistic interpretability reveals that safety-critical behaviors (e.g., alignment, jailbreak, backdoor) in Large Language Models (LLMs) are grounded in specialized functional components. However, existing safety attribution methods struggle with generalization and reliability due to their reliance on heuristic, domain-specific metrics and search algorithms. To address this, we propose \ourmethod, a unified safety interpretability framework that identifies functionally complete safety circuits in LLMs via optimization. Unlike methods focusing on isolated heads or neurons, \ourmethod introduces differentiable binary masks to extract multi-granular circuits through gradient descent on safety datasets, while integrates Safety Circuit Tuning to utilize these sparse circuits for efficient safety fine-tuning. We validate \ourmethod in two key scenarios in LLM safety: \textbf{(1) backdoor attacks}, identifying a backdoor circuit with 0.42\% sparsity, whose ablation eradicates the Attack Success Rate (ASR) from 100\% $\to$ 0.4\% while retaining over 99\% general utility; \textbf{(2) safety alignment}, localizing an alignment circuit with 3.03\% heads and 0.79\% neurons, whose removal spikes ASR from 0.8\% $\to$ 96.9\%, whereas excluding this circuit during helpfulness fine-tuning maintains 96.5\% safety retention.
cs.LG↗
ASK: Adaptive Self-improving Knowledge Framework for Audio Text Retrieval
The dominant paradigm for Audio-Text Retrieval (ATR) relies on dual-encoder architectures optimized via mini-batch contrastive learning. However, restricting optimization to local in-batch samples creates a fundamental limitation we term the Gradient Locality Bottleneck (GLB), which prevents the resolution of acoustic ambiguities and hinders the learning of rare long-tail concepts. While external knowledge injection can break this bottleneck, it often triggers a problem called Representation-Drift Mismatch (RDM), where a static knowledge base becomes misaligned with evolving encoders, degrading guidance into noise. To address these intertwined challenges, we propose the Adaptive Self-improving Knowledge (ASK) framework. ASK breaks the GLB via multi-grained knowledge injection and mitigates RDM through a dynamic refinement strategy that synchronizes the knowledge base with the model. Additionally, an adaptive reliability weighting scheme is employed to filter retrieval noise based on cross-modal consistency. Extensive experiments across multiple benchmarks demonstrate that ASK consistently achieves new state-of-the-art performance across various backbones.
eess.AS↗
High Precision Measurements Using High Frequency Signals
Generalized lock-in amplifiers use digital cavities with Q-factors as high as 5X10^8. In this letter, we show that generalized lock-in amplifiers can be used to analyze microwave (giga-hertz) signals with a precision of few tens of hertz. We propose that the physical changes in the medium of propagation can be measured precisely by the ultra-high precision measurement of the signal. We provide evidence to our proposition by verifying the Newton's law of cooling by measuring the effect of change in temperature on the phase and amplitude of the signals propagating through two calibrated cables. The technique could be used to precisely measure different physical properties of the propagation medium, for example length, resistance, etc. Real time implementation of the technique can open up new methodologies of in-situ virtual metrology in material design.
physics.ins-det↗
Generalized lock-in amplifier for precision measurement of high frequency signals
We herein formulate the concept of a generalized lock-in amplifier for the precision measurement of high frequency signals based on digital cavities. Accurate measurement of signals higher than 200 MHz using the generalized lock-in is demonstrated. The technique is compared with a traditional lock-in and its advantages and limitations are discussed. We also briefly point out how the generalized lock-in can be used for precision measurement of giga-hertz signals by using parallel processing of the digitized signals.
physics.ins-det↗