Multimodal Emotion Recognition (MER) systems often suffer from missing modalities in real-world scenarios. Existing methods usually generate, align, or distill missing modalities as a whole, overlooking the heterogeneous nature of the information carried by each modality. Such holistic treatment mixes inferable shared semantics with uncertain modality-specific details, yielding unstable representations and degrading robustness. To address this issue, we propose the Primitive Memory Distillation (PriMD) framework. Unlike existing methods, PriMD takes an intra-modal perspective and focuses on how different types of information within a modality differ in recoverability within each modality. PriMD first disentangles cross-modal shared semantics from modality-specific representations, and then discretizes the latter into learnable semantic primitives to construct modality-specific memory banks. When modalities are missing, PriMD is a teacher–student framework that the student model uses the shared semantics of available modalities as queries to dynamically retrieve primitives. It compensates for missing modality-specific information within a constrained memory space and aligns with the teacher model. Extensive experiments on IEMOCAP, CMU-MOSI, and CMU-MOSEI demonstrate that PriMD achieves state-of-the-art performance and consistently stronger robustness across a wide range of missing-modality settings, while mitigating the instability caused by holistic feature inference. Our code and project website are available at https://github.com/JiaqiZhang-Sengoku/PriMD and https://jiaqizhang-sengoku.github.io/PriMD/, respectively.

Long-context inference in large language models is limited by the linear memory growth of the key-value (KV) cache. Existing eviction methods typically retain tokens with high accumulated attention or remove semantic redundancy, but may discard low-frequency tokens that connect distant dependencies. We propose CentralKV, a training-free framework that models attention as a dynamic sparse causal graph. Its Stream-Influence algorithm updates token centrality online and preserves structurally important bridge tokens during KV compression. The attention graph is constructed directly from the attention operation without separate similarity computation. Experiments on LongBench show that CentralKV improves over H2O, SnapKV, and GraphKV under strict physical memory constraints, particularly on reasoning-intensive tasks, with minimal latency overhead.

Group-relative policy optimization (GRPO, DAPO, ARPO) has become the dominant recipe for RL fine-tuning of reasoning LLMs, yet in the resource-constrained regime most practitioners train in (a single GPU, small batches) it consistently suffers from rapid entropy collapse that kills exploration, together with sparse credit assignment in long rollouts, which is most extreme under multi-turn tool use but already present in long single-turn chain-of-thought. We address both failure modes with JANUSPO, a drop-in GRPO modification that introduces asymmetric PPO clipping 𝜖₊<𝜖_- to tolerate larger downward than upward updates, pairs it with an asymmetric KL schedule 𝛽_pos<𝛽_neg that doubles as a free entropy regulator without any explicit upper bound, and redistributes the terminal reward through a turn-weighted decomposition w_t(n) that generalises naturally from single-turn CoT to multi-turn rollouts. On Qwen2.5-3B-Instruct / GSM8K, JANUSPO improves from 74.8% to 84.3±0.2% (mean±std over 2 seeds) across 14 controlled ablations, and the same recipe lifts 7B from 85.3% to 93.6% with no hyper-parameter retuning. On multi-turn DeepSearch (103 GAIA + 100 HLE), SFT-warmup-then-JANUSPO is the only RL recipe whose paired comparison against SFT survives best@8 outlier verification, achieving strict dominance on all four reward and F1 columns while a matched-init GRPO baseline’s apparent F1 wins collapse to 0%. Beyond this main result, we document a systematic negative finding: SFT warm-up is a necessary precondition for GRPO-style multi-turn RL on format-unaware base models.