多模态生成与编辑

Long-form text-to-speech must retain requested words over extended generations, yet sentence-level training and evaluation can hide omissions and early stops. We introduce Balalaika-Longform, an open Russian corpus of 189 hours in continuous units of 30 seconds to 15 minutes. Long units and matched short windows support fine-tuning comparisons on the same source recordings, and the accompanying evaluation retains every synthesis attempt. We fine-tune CosyVoice3, Qwen3-TTS, VoxCPM2 and F5-TTS on either view and test continuous synthesis on 50 text-voice pairs with voices unseen in fine-tuning, at about 75, 300 and 1,200 words. At 1,200 words, CosyVoice3 WER falls from 99.9% after short-window fine-tuning to 47.3% with long targets, and to 16.6% with punctuated, format-matched transcripts; paired intervals favor long targets for Qwen3-TTS and VoxCPM2, and by a small margin for F5-TTS, whose absolute WER stays above 90%. The results isolate the effect of training sequence length under a fixed continuous-generation protocol.

Dance-to-music (D2M) generation aims to synthesize musically plausible soundtracks whose temporal structure aligns with a given dance performance. Representative D2M methods do not explicitly distinguish motion cues across body parts and frequency bands, while standard flow matching lacks a dedicated objective for supervising local music-latent dynamics. To address these issues, we propose Dyna2Music, a latent flow-matching framework that combines structured motion conditioning with explicit supervision of music-latent dynamics. An empirical analysis on AIST++ quantifies how spatial partitioning and frequency separation affect raw music-to-kinematic beat alignment and motion-reference density, informing the conditioning design. Accordingly, Dyna2Music decomposes joint velocities into slow and fast components and hierarchically fuses the resulting part-wise motion energy with pretrained joint features to condition music generation. To complement this representation, we introduce latent dynamics consistency (LDC), an auxiliary objective that matches adjacent-frame change magnitudes between a single-step clean-latent estimate and the paired reference. LDC makes local music-latent variation an explicit training target without adding trainable parameters or inference computation. Dyna2Music supports variable-length music generation, and experiments on AIST++ and TikTok demonstrate improved rhythmic alignment and audio quality over representative D2M baselines.

Modern neural text-to-speech (TTS) systems achieve remarkable acoustic fidelity but act as black boxes, offering little interpretable control over the vocal tract filter. We propose a controllable source-filter TTS architecture grounded in articulatory kinematics. An Acoustic-to-Articulatory Inversion (AAI) model, enhanced by large-scale pretrained representations, generates kinematic pseudo-trajectories for a large TTS corpus. These trajectories condition the filter response, while predicted pitch and energy contours parameterise the glottal source. Source and filter are predicted independently, and the source is refined by an Optimal Transport Conditional Flow Matching (OT-CFM) module before recombination into the final spectrogram. Our model achieves intelligibility and naturalness competitive with similarly sized baselines, with only a modest spectral fidelity cost in exchange for explicit control. Evaluations reveal clear source-filter disentanglement, enabling stable prosodic scaling and cross-speaker source/filter recombination, where F0 remains tied to the source speaker while vocal-tract characteristics follow the filter speaker. Finally, direct manipulation of articulatory trajectories enables fine-grained control, such as accent modification, offering a new direction for interpretable speech synthesis. Audio samples: https://coding-phoenix-12.github.io/ArticulatorySFTTS/

Recent LLM-based approaches to control either invoke a language model to select actions or synthesize world models that require planning at every decision, introducing latency that can limit real-time use. We introduce Code to Control, an approach that synthesizes Python controllers which execute directly as policies. Code to Control separates program structure from parameters. An LLM synthesizes the controller structure, while derivative-free search fits its parameters for continuous control using feedback from the environment. Once learned, the resulting controllers require neither LLM inference nor planning at decision time, enabling real-time gameplay and, under our timing protocol, faster action selection than a PPO policy. Across a suite of Atari games, Flappy Bird, and MuJoCo tasks, Code to Control outperforms planning-based program synthesis methods, remains competitive with deep reinforcement learning while using fewer environment interactions, transfers across substantial changes in environment dynamics, and scales to complex locomotion tasks.

Sparse-view 3D Gaussian Splatting is prone to overfitting because limited observations leave many Gaussian primitives weakly constrained, yet their contributions are still accumulated through alpha blending. Without uncertainty estimation, the renderer cannot distinguish unreliable primitives from well-constrained ones, allowing their erroneous contributions to corrupt novel-view synthesis. We introduce UGOD, an uncertainty-guided framework that estimates a view-dependent uncertainty score for each Gaussian and uses it to regulate its rendering contribution. A lightweight uncertainty head conditioned on Gaussian attributes and viewing direction predicts this score, which then drives a differentiable opacity-modulation mechanism that attenuates high-uncertainty primitives before compositing. During training, a detached soft-dropout branch applies an uncertainty-controlled continuous keep mask to discourage the model from relying on poorly constrained Gaussians and thereby reduce overfitting. Crucially, detaching the uncertainty score prevents gradients from this stochastic regulariser from biasing or collapsing the uncertainty prediction. Experiments on Mip-NeRF~360 and LLFF show that UGOD improves sparse-view novel-view synthesis while producing more compact Gaussian representations than the compared methods. These results demonstrate that Gaussian uncertainty provides an effective rendering-time control for sparse-view reconstruction.

3D Gaussian Splatting (3DGS) enables real-time novel view synthesis, but existing general-purpose acceleration methods suffer severe rendering quality degradation when extended to more complex, large-scale scenes. To address this issue, we propose EffGS, a more general acceleration framework that improves training and rendering efficiency while maintaining reconstruction quality comparable to or better than vanilla 3DGS across bounded and large-scale scenes. EffGS combines frequency-aware guidance, localized density control, and adaptive primitive scale modulation. First, an importance scoring mechanism combines pixel-wise reconstruction errors with a difference-of-Gaussians mask scheduled over training to provide stage-dependent spatial guidance. Second, localized densification and pruning restricts density modifications to Gaussians with valid projected footprints in the sampled views. Third, learnable per-Gaussian scale modulation adjusts effective primitive extent during optimization while retaining the Compact Box rasterization rule. Extensive experiments on bounded and large-scale scene datasets demonstrate a favorable balance between reconstruction quality, training time, and primitive count. Component ablations and matched-primitive-budget comparisons further support the effectiveness of the framework.

Compositional 3D scene generation aims to recover complete 3D object shapes and their spatial arrangement from visual observations. Recent image-conditioned 3D generators provide strong priors for producing high-quality object geometry, making the generation of complex scenes increasingly practical. A central challenge is therefore to spatially organize these generated assets into a globally coherent scene while remaining consistent with multi-view observations. Existing approaches either entangle scene layout with object generation or separately estimate spatial placement from view-specific observations, where pose hypotheses may remain ambiguous and inconsistent across views, often resulting in an incoherent object-camera soup. We introduce SPOON, a framework that reformulates multi-view compositional 3D generation as scene-level, geometry-grounded pose reasoning. Rather than treating view-specific object pose hypotheses independently, SPOON coordinates them using reconstruction-derived multi-view geometry through a Guide-Route-Reconcile paradigm. This progressively organizes object poses and camera configurations into a coherent scene-level spatial arrangement. Extensive experiments on ARSG-110K and MIDI-3D-Front demonstrate consistent improvements in object placement and scene composition across varying numbers of input views. On ARSG-110K, SPOON reduces scene-level and object-level Chamfer distances by 12.7% and 17.7%, respectively, compared with a strong baseline.

3D visuomotor policies provide a strong foundation for spatially precise manipulation, yet current text-to-3D policies struggle to follow unseen fine-grained behavioral specifications beyond those covered by demonstrations. We study this challenge as unseen specification generalization, where language specifies behaviorally significant variations, such as target position, displacement, or articulated state, that are absent from policy training. We find that pretrained language representations and conventional global behavior-language alignment capture coarse task semantics but often blur nearby specifications that require distinct behaviors. We introduce T3DP, a Text-to-3D Policy framework for fine-grained language-behavior alignment. Rather than compressing each instruction and demonstration into a single global embedding, T3DP preserves their local structures and establishes bidirectional token-level correspondence between linguistic elements and behavioral segments. This directly grounds subtle linguistic variations in the behavior components they affect, preventing closely related specifications from collapsing in the representation space. The resulting specification-sensitive language representation conditions a point-cloud-based 3D diffusion policy, enabling more precise control over unseen behavioral specifications without modifying the underlying policy architecture. Across Meta-World, ManiSkill, and RoboTwin, T3DP improves average held-out-specification success over global language-behavior alignment by +11.0-14.2 points, with gains on all 15 task families; on real-robot tasks, it further raises average success from 47.5% to 65.0% (+17.5 points). Representation and action-probe analyses show that fine-grained alignment better preserves specification geometry and action-relevant variation, linking local behavior grounding to downstream control.

Sparse autoencoders (SAEs) reveal visual structure in diffusion transformers (DiTs), but interpreting a feature does not establish whether it can be used to control generation. We introduce D-Scope (Diffusion Scope), a framework that connects feature interpretation to generation control through shared visual evidence. D-Scope aggregates SigLIP~2 embeddings of highly activating image patches into visual centroids. Matching target text descriptions against these visual centroids in the shared image-text embedding space then enables retrieval of individual features without per-feature text annotations. The underlying patches provide evidence for inspecting each selection, while spatially masked interventions test the corresponding decoder direction at varying strengths under fixed generation conditions. We characterize 150 SAEs across two model families and five layers, and introduce a benchmark of 100 target concepts with ten contexts each spanning under-specified and explicit-conflict conditions. Our empirical results show that high reconstruction fidelity can coexist with low dictionary utilization and limited visual-evidence coverage. Under per-case best-of-sweep strength selection, contrastive retrieval yields larger mean regional SigLIP~2 gains than direct retrieval across the tested steering configurations, without consistently improving outside-region preservation. D-Scope provides an inspectable framework for evaluating sparse DiT features through their visual evidence and the effects of their decoder directions on generation. The demo is available at https://jiahaozhang-public.github.io/d-scope/.

Reconstructing dynamic driving scenes from recorded sensor data supports closed-loop evaluation of autonomous driving systems by synthesizing observations beyond the original trajectory. Unlike cameras and LiDAR, radar measures radial velocity directly through Doppler. Yet existing radar novel-view synthesis fails to exploit this capability: methods addressing dynamic scenes reconstruct only range-azimuth tensors, while methods that render Doppler assume static scenes. Moreover, because radar processing spreads each reflection across multiple bins, existing representations absorb this spread into scene geometry, causing it to render incorrectly when the viewpoint moves. We present DyRAD, which models dynamic driving scenes using static background reflectors and motion-tracked dynamic point reflectors to render complete range-azimuth-Doppler (RAD) tensors. Reflector velocities are derived from object tracks and projected onto the line of sight, making Doppler both a rendered output and supervision for those tracks. Crucially, we render reflectors through a fixed analytic point-spread function (PSF) derived from the radar's signal-processing chain, preventing sensor-induced spread from being baked into the scene representation. Beyond improving scene reconstruction, this separation also enables zero-shot sensor-configuration transfer, allowing the same reconstructed scene to be rendered under different radar specifications without refitting. We evaluate DyRAD on RADIal, Boreas, and a synthetic benchmark across both on-path poses and displaced viewpoints untested by prior work. On RADIal, DyRAD recovers radar detections in 90.7% of reference-detected objects, compared with 26.9% for the strongest baseline.

Video world models are expected to predict how the physical world evolves, yet they often produce visually plausible videos that violate basic physical principles. Existing approaches commonly assume that natural language is insufficient to represent the physical knowledge required for reliable generation, and therefore introduce additional visual, latent, numerical, or planning-based signals. We revisit this assumption and introduce Physis-Lang, a self-evolving framework that treats physical language as a shared and optimizable representation across data curation, model training, and video generation. Physis-Lang represents physical processes through language that describes their relevant entities, causes, interactions, governing principles, temporal evolution, and effects. To improve this representation, we construct PhysCapBench, which decomposes physical processes into atomic assertions and evaluates captions using recall and precision. An agentic loop iteratively analyzes assertion-level errors and refines the instruction used to produce physical captions. Physis-Lang further converts model deficiencies into textual descriptions and uses language-guided retrieval to identify visually diverse videos that cover missing physical processes. Experiments on four widely used physical video benchmarks with Wan and Cosmos backbones demonstrate consistent improvements in physical plausibility. Notably, starting from open-source Cosmos3-Nano backbones, our Physis-Lang-enhanced models surpass the leading proprietary Veo 3.1 model.

Can a text-to-3D leaderboard change when every generated scene stays fixed? We audit this question for rendered-image evaluation, where camera settings and caption wording become part of the measurement protocol. Across 300 frozen scenes from six generators, we vary eight render and caption factors for 19 alignment evaluators plus one perceptual-quality control, then test four targeted scene degradations. Peak configuration variance exceeds between-generator variance for 17/19 alignment evaluators, with prompt-bootstrap lower bounds above 1 for 11/19. Rankings are more stable than scores, yet 18/19 evaluators change their point-estimate winner under some configuration. Pairwise protocol margin envelopes show which comparisons keep their direction across the tested settings. Selected pairs have opposite pointwise intervals, but no reversal survives simultaneous inference over the full search. Thus the observed winner changes are descriptive, not confirmed changes in generator superiority. Sensitivity remains separate: no evaluator, even the prompt-free control, exceeds 67% tie-adjusted directional discrimination on layout scrambling, which is diagnostic rather than human-validated ground truth. The audit separates score stability, decision uncertainty, and targeted sensitivity, and recommends reporting (generator, score, card ID) with protocol-dependent comparisons and selection-aware uncertainty.

Streaming world models should render a place consistently across repeated visits. Directly supervising such revisits requires training samples that capture both visits, often spanning minutes. Yet dense attention over the full span incurs quadratic costs, making long-span supervision expensive. Memorizon breaks this coupling: long spans are needed for supervision, but not for attention, since the two visits can share a forward pass without including every intervening frame. A training sample covers a span of any length but is scored only on its last $k$ chunks. Instead of tokenizing the history before them, each scored chunk retrieves its own top-$K$ latents by camera co-visibility, and the union of these requests forms a shared bank. The bank is bounded by $kK$, so the sequence stays bounded however long the span; at the shortest span the recipe is exactly conventional training. Adding the bank raises the cost of a step once; beyond that, a longer span costs little, and going from 100 to 400 s adds 12% to the step time. Against a sliding-window baseline, retrieval raises revisit consistency on every split, and a span long enough to reach the first visit of each return adds a further 24% to 30%, at some cost in image quality; beyond that span, more length no longer helps. Filling the bank from another episode lowers revisit correlation by 83%, so the model uses what it retrieves. Project page: https://tingtingliao.github.io/memorizon

Text-to-music generation models are trained on massive music collections, creating a growing need for data attribution methods that can quantify the contribution of individual training samples. However, existing attribution methods are difficult to rigorously evaluate due to the lack of reliable ground truth, making it challenging to reliably assess their actual effectiveness. To address this gap, we introduce TrueMuse, a controlled dataset and benchmark for text-to-music data attribution. TrueMuse is constructed by fine-tuning three diffusion-based text-to-music models on carefully curated attribution samples, whose known inclusion in fine-tuning provides controlled attribution targets for evaluation. The benchmark covers four attribution settings, spanning melodic structure, timbral characteristics, artist-level stylistic signatures, and genre-level shared patterns, and includes 133 attributes, 648 fine-tuned models, and 95,456 generated samples across two prompt types. Using TrueMuse, we systematically evaluate existing black-box attribution methods along four dimensions: fine-tuning improvement, prompt-type difficulty, multi-task training, and fine-tuning data size. Our results show that attribution remains challenging, with existing methods exhibiting substantial variation across evaluation settings, highlighting the need for more reliable and generalizable attribution methods for text-to-music generation. Code and Dataset will be released upon acceptance.

In this paper, we propose Gestalt, a new paradigm of large multimodal model built around multimodal interplay. Despite rapid advances, large multimodal models are reaching a bottleneck: existing approaches focus primarily on accommodating additional modalities while overlooking the distinct characteristics of each modality and the relations among them. Motivated by the multistage property of human multisensory perception, we propose a multimodal interplay pyramid that organizes multimodal modeling as a progression from modality-specific processing, through cross-modal alignment, to deeper multimodal integration. Guided by this pyramid, Gestalt adopts a unified discrete diffusion framework and an interplay-partitioned architecture, with learnable interplay tokens mediating cross-modal exchange and integration. The pyramid also structures its data organization and training strategy. Strong performance across image generation, multimodal understanding, and text-only evaluation shows that Gestalt significantly improves cross-modal integration while preserving modality-specific information, effectively harnessing the strengths of diffusion-based multimodal models and offering a promising path toward unified multimodal intelligence. Project page: https://GeWu-Lab.github.io/Gestalt.

World-model planners typically scale outward by rolling farther, sampling more trajectories, or optimizing longer, while assigning the same computation to every imagined transition. We show that making every transition uniformly deeper wastes computation and can degrade planning because useful refinement is concentrated at a small set of decision-critical events. We introduce DeepJEPA, a weight-tied joint-embedding predictive world model that treats transition depth as an inner test-time scaling axis and learns when another recurrent update is worth computing for each candidate and rollout step. Across five visual-control settings, DeepJEPA improves or matches the strongest fixed-depth planner while averaging only 1.00-1.26 updates per transition. Its additional computation concentrates at contact onset and sustained object interaction, where latent corrections can change which candidates enter the planner's elite set and which action is selected. Representation probes further show that improved planning does not require uniformly better object-state decodability. DeepJEPA therefore reframes world-model scaling as a problem of allocating internal computation where it can change the planner's decision: think deeper at decision-critical transitions instead of making every rollout uniformly deeper or longer.

Controlling an agent with vision requires being able to separate useful information from irrelevant background information. JEPA-style latent world models seem like a natural approach for this, as they do not perform pixel-level reconstruction; however, they are still sensitive to these distractor signals and experience latent collapse. In this work, we introduce Controllability Factorized JEPA (CF-JEPA), a JEPA-style world model which splits the latent space into controllable and uncontrollable subspaces. This factorization allows us to capture all the distractor information into the uncontrollable region, while we use the control-relevant latent information for our task. With this, we show comparable performance across 2D and 3D control tasks under nominal conditions and improved performance under distracted conditions, where CF-JEPA is the only model that does not experience latent collapse. We also validate our model under distracted conditions for a simulated robot task, highlighting the practical application of such a scheme.

Dubbing quality control requires a reference-free judge that can determine whether a candidate text line matches a speaker's visible articulation in both content and timing, using only silent video and text because dubbed audio may not yet exist. Existing visual speech recognizers and video-language models are poorly suited to this setting: even when fine-tuned to recover spoken content from lip motion, they remain largely insensitive to temporal errors. We introduce $Align Then Reason$ (ATR), a multilingual lip-sync judge that first establishes a monotonic alignment between frame-level lip representations and the phonetic units of the candidate line, then reasons over this alignment to make the final judgment. An alignment scorer provides the LLM with both local evidence for each phonetic unit and a calibrated global alignment score, enabling it to reason jointly about content and timing. On a seven-language benchmark, our method improves mean AUC over the corresponding Qwen3.5 SFT baselines by 59.4%, 50.2%, and 50.8% with 2B, 4B, and 9B reasoners, respectively. The gains generalize across LLM families, reaching mean AUC improvements of 45.9% and 46.6% over the best baseline for LLaMA-3.1-8B and Mistral-7B, respectively. They also transfer across datasets to three unseen MuAViC languages. Furthermore, we evaluate on two downstream tasks built from real dubbing lines. On dub-line reranking, ATR-9B outperforms the best lip-reading baseline by 52.0%, while on script-to-clip assignment, ATR-9B improves over the best lip-reading baseline by 17.7%.

World models enable agents to plan by predicting future states of the environment, but their predictions can become unreliable when test-time dynamics differ from those seen during training. We present JEPA-TTT, which adapts the latent dynamics predictor of a pretrained action-conditioned Joint-Embedding Predictive Architecture world model throughout test time. Self-supervised updates accumulate across episodes, while the visual encoder and reward head remain fixed, preserving the pretrained representation and task objective. Planning requires neither a goal image nor online environment reward. JEPA-TTT uses dense replay, which forms prediction windows at every temporal offset, retains them in a growing buffer, and samples minibatches from that buffer for predictor updates. Across eight dynamics shifts in four continuous-control environments, JEPA-TTT improves planning on every shift. After 500 test-time episodes, it reduces autoregressive latent prediction error by 83% on average and improves planning performance by 153% over the frozen JEPA world model. These results show that persistent self-supervised test-time training can adapt a pretrained latent world model under changed dynamics.

ARC-AGI-3 evaluates agents in interactive environments whose rules and objectives must be inferred from observation. We present Kepler, an open-source harness that represents hypotheses as executable world models and validates them through retrospective transition checks and conditional prediction checks. Under one frozen Claude Opus 5 configuration, Kepler obtained a server-verified 100.00 RHAE on all 25 public games, with no per-game model selection or score-conditioned reruns. On 181 of 183 completed levels, the final Opus attempt used no more actions than the corresponding median-human baseline. The retained board runs used 8,256 environment actions, of which 7,292 occurred in scored levels. Retained local provider-session records yield 858.0 million tokens, 97.37% cache reads, and a \$777.72 cost at September 1, 2026 API list-equivalent rates. We also report three evaluation failures: source-code leakage that produced an invalid perfect run, agents reconstructing a removed harness in a control condition, and autonomous repair masking a broken planner. A single-game observation case study showed that animation frames contained task-relevant information absent from settled text grids. Across the final Claude Opus 5 and GPT-5.6 Sol boards, 48 of 50 game-model cells reached 100. These results indicate that public-set score alone has limited discriminative value and motivate first-attempt, cost-conditioned, and verification-aware reporting.

Modern text-to-image (T2I) systems can be improved without modifying generator parameters by adapting the external system around frozen generators. However, existing approaches typically optimize a predefined dimension, such as prompts, routing, or workflows, restricting the space in which generation failures can be corrected. Allowing multiple generator-external responsibilities to evolve provides a broader adaptation space, but introduces a new challenge: visual feedback reveals what failed, but not where persistent evolution should occur or how this space should be explored efficiently. We introduce EvoGen-Harness, a generator-agnostic framework for multi-responsibility image-generation harness evolution, together with Trace (Trajectory-Relative Attribution and Coordinated Evolution). Trace aggregates evidence across stochastic executions, uses failure attribution as a search prior to focus candidate updates, and progressively re-attributes residual failures to coordinate evolution across responsibilities, while No-Patch and held-out validation prevent unnecessary or harmful updates. Across GenEval2, T2I-CompBench++, and WISE, EvoGen-Harness improves over the strongest evaluated baselines by +0.2633, +0.0720, and +0.0752, respectively, while achieving 87.9-91.4% attribution recall, 94.8% No-Patch accuracy, and only 1.9% regression. These results demonstrate that attribution-guided multi-responsibility evolution can substantially enhance frozen T2I systems beyond single-dimension adaptation.

Recent video-based world models pair the scalability of autoregressive (AR) prediction with the visual quality of diffusion models. The choice of scene tokenizer is paramount for the optimal performance of each of these, both in terms of fidelity and semantics. Flexible-length, coarse-to-fine tokenizers yield exactly that: the first coarse tokens carry the clip's global semantics while later tokens further specify details. Existing flexible tokenizers only apply a representation-alignment (REPA) loss on early decoder hidden states, a target the decoder can partly meet from its noised input instead. We introduce SemanTok, a flexible video tokenizer that feeds frozen DINO features into its encoder and adds lightweight heads that reconstruct them from each retained token prefix alone. SemanTok achieves high semantic alignment and video fidelity at every AR model size: a 201M SemanTok AR model matches or beats a VideoFlexTok AR model $3.4\times$ its size, and larger SemanTok AR models further improve fidelity. It keeps semantic alignment on out-of-distribution classes and gives the decoder higher semantic alignment at every noise level, including pure noise. It performs well in both reconstruction and generation, and its short token prefixes are cheaper to predict and give better generation fidelity, with pixel detail deferred to later tokens.

We present PLACE, a method that extends the pretrained any-to-audio model AudioX for binaural generation from arbitrary combinations of text, video, and optional audio prompts. PLACE augments video conditioning with Perception Encoder Core features, aligns text and video representations to derive spatial cues, and applies a conditioning-dependent low-rank transformation to the generated latent. The adapter is supervised via decoded-audio interaural level and time difference objectives. Trained on MRSAudio, PLACE improves most metrics over ViSAGe on FAIR-Play and achieves an improved SpatialCLAP score over SpatialSonic on the BEWO-1M Single Static test split. Listener evaluations favor PLACE for video-to-audio and out-of-distribution text-to-audio generation, demonstrating flexible multimodal control and improved spatial consistency.

Personalized image generation has remained narrowly focused on conditional synthesis from curated visual exemplars, rather than capturing who a user is. In practice, however, a user's personal context is much richer, comprising reviews, posts, images, captions, and metadata accumulated over time. A truly personalized generator should leverage this history to produce images aligned with the user's lifestyle and aesthetic preferences. To this end, we introduce the first unified benchmark for personalized image generation from user histories. The benchmark comprises two complementary tasks and a multi-axis evaluation protocol that assesses target fidelity, visual quality, user distinguishability, semantic alignment with the user's history, and task-specific utility. Grounded in real-world e-commerce and social media settings, the benchmark includes: (1) Personalized Scene Generation, which places a given object in a scene that reflects a user's preferences and lifestyle, motivated by personalized product presentation; and (2) Personalized Creative Generation, which generates a novel image on a specified topic that is faithful to a user's aesthetic and visual identity, motivated by social media content creation. We further propose PEARL, which couples a multimodal reasoner with a frozen image generator in an interleaved reason-reflect loop optimized with differential data reward. Across both tasks, PEARL outperforms strong baselines, achieving an average improvement of 15% across personalization metrics.

Recent advances in distillation and flow-map models have enabled deterministic one- or few-step generation for high-quality data, facilitating a new branch of reward alignment approaches that directly optimize the initial noise from a Gaussian distribution. However, most existing initial-noise optimization methods rely on first-order gradient information, which is either inapplicable or suffers from instability and inefficiency in black-box reward scenarios. Here, we introduce ZeNOVA, a stable and efficient initial noise alignment method in a gradient-free manner. Specifically, we address existing algorithms' major challenge in black-box scenarios through annealed soft-value guidance, manifold-constrained hyperspherical Langevin dynamics, and Metropolis-Hastings jumping. Extensive experiments on image and video generative models show that ZeNOVA outperforms all evaluated zeroth-order baselines by optimizing the initial noise toward higher rewards substantially more stably while exploiting the geometry of the Gaussian prior, demonstrating its practical applicability to various black-box reward alignment.

Recent joint audio-video generative models can synthesize realistic videos with synchronized sound, but typically generate audio as a single mixed track. This limits source-level control and differs from practical audiovisual workflows, where speech, music, sound effects, and ambient sounds are represented as separate editable tracks. We introduce Soundwich, a training-free framework that transforms a frozen joint audio-video flow-matching model into a generator of multiple synchronized, independently editable audio stems coupled to a shared video. Soundwich generates separate audio stems with explicit control over their temporal activity. To keep separately generated sounds coherent, we introduce a shared scene representation that communicates global audiovisual context across stems while preserving their source-level separation. We further route cross-modal interactions between each audio stem and its corresponding visual source, improving audiovisual consistency. The resulting stems remain synchronized with the video and can be independently retimed, muted, replaced, or remixed. Experiments and human evaluations show improved temporal control, source separation, and naturalness, while enabling flexible source-level editing within coherent audiovisual generation. Code is available at https://github.com/CodyNing/Soundwich.

Diffusion models with prompt and reference image-guided editing have seen rapid progress, yet they remain too coarse for pixel-level control. One promising direction is to incorporate a soft mask that specifies spatially varying edit strengths but training such fine-grained control demands expensive pixel-wise annotations, while existing zero-shot methods often yield unsatisfactory results. We introduce SoftPaint, a new zero-shot sampling method that leverages soft masks to enable a continuous spectrum of edits, from fully preserving the original content to completely re-synthesizing the masked region. Going beyond zero-shot inpainting methods, we design a Langevin-iteration-based sampler that respects per-pixel soft mask strengths, which applies universally to image and video diffusion models, enabling tasks such as video editing. The method is gradient-free, memory-efficient, and achieves smooth, pixel-level edits across multiple image and video backbones.

Inference-time steering enables pretrained diffusion models to satisfy new constraints without full retraining. However, specificity-aware generation is difficult: repelling samples from a negative reference distribution can also erode the positive distribution where the two overlap. The key challenge is to suppress negative mass while minimally distorting the positive distribution. We address this problem by formulating specificity-aware steering as a target-design problem and deriving a target distribution from an overlap-based objective. The resulting target keeps the desired reference distribution only in regions where it is sufficiently preferred over the undesired reference distribution, giving a likelihood-ratio interpretation of specificity. To sample from the corresponding time-dependent target path, we develop a Sequential Monte Carlo sampler with a variance-minimized local proposal. We further introduce a practical fixed-noise optimization procedure with the Jacobian--vector products with the desired and undesired score fields. Experiments on synthetic task, class-contrastive generation, text-to-image tasks and peptide-MHC (p-MHC) binder show that the proposed method suppresses undesired regions more effectively, reduces mode shift, and improves sampling stability by decreasing the SMC weight collapse compared with negative-guidance baselines. Code is available at: https://github.com/WangLuran/Specificity-Aware-Diffusion-Steering

Multimodal encoders such as CLIP underlie many downstream systems, but their web-scale training data embed harmful associations that safety alignment must suppress without unnecessarily changing benign representations. Because ethical and practical constraints prevent collecting real unsafe content at scale, existing datasets pair safe real samples with generated counterparts, but label every generated sample unsafe, even when one modality is individually safe. To address this, we introduce ShieldCLIP, the first framework to condition safety alignment on the observed safety state of each modality rather than the origin of a sample, preserving safe content while redirecting only what is unsafe. We also introduce ViSUv2, a 195k-quadruplet dataset with independent per-modality safety labels across 578 concepts and 28 categories. Using these labels, ShieldCLIP defines a four-way conditional objective beyond pair-level supervision: safe content is anchored, unsafe modalities are redirected to their safe counterparts, mixed pairs update only the unsafe branch, and coherence is enforced when both are unsafe. We evaluate ShieldCLIP on cross-modal retrieval, text-to-image generation with Stable Diffusion v1.4 and SDXL, and image-to-text generation with LLaVA. Across these settings, ShieldCLIP consistently reduces harmful outputs over prior safety-aligned encoders and strong mitigation baselines, while preserving the utility of the original embedding space. Extensive ablation studies further show that both modality-specific supervision and the selective alignment objective contribute to these gains. Source code, trained models, and ViSUv2 (under a controlled-access protocol) will be made publicly available at https://aimagelab.github.io/ShieldCLIP/.

Modern multimodal models bring generation and understanding into a single unified system, which enables them to provide and learn from their own feedback. Motivated by this unified capacity, we introduce UniEvo-VL, a self-evolving framework for multimodal models to learn from this constructive self-correction feedback during test-time compute. Instead of relying on a separate, often larger, teacher, we leverage their self-critiques as privileged information and ask a single multimodal model to act as both teacher and student with different contexts. The student only sees the vanilla question, while the teacher conditions on the privileged critique. Then training minimizes the per-state divergence between their denoising diffusion distributions over the student's own sampling trajectories. Experiments demonstrate that UniEvo-VL improves the image generation capabilities of multimodal models, while maintaining their sensitivity to additional reflection information. Specifically, we build on top of the open-source Qwen-image-2512 and observe a significant performance gain from 0.747 to 0.808 on GenEval and from 32.97 to 35.53 on GenEval2 Soft-TIFA. Moreover, attempts with more powerful external critics (e.g., GPT5.6-Luna) show that multimodal models with strong judge capabilities can anticipate a higher self-evolving ceiling. Last but not least, mixed text-rendering outcomes show that our self-improvements may not be uniform across different tasks. Our study aims to shed light on the current hot recursive self-improvement research line to enhance the user experience when using multimodal models without external supervision or guidance.