Search

Large language models (LLMs) and vision-language models (VLMs) have demonstrated remarkable performance across a wide range of tasks and domains. Despite this promise, spatial understanding and reasoning -- a fundamental component of human cognition -- remains under-explored. We develop novel benchmarks that cover diverse aspects of spatial reasoning such as relationship understanding, navigation, and counting. We conduct a comprehensive evaluation of competitive language and vision-language models. Our findings reveal several counter-intuitive insights that have been overlooked in the literature: (1) Spatial reasoning poses significant challenges where competitive models can fall behind random guessing; (2) Despite additional visual input, VLMs often under-perform compared to their LLM counterparts; (3) When both textual and visual information is available, multi-modal language models become less reliant on visual information if sufficient textual clues are provided. Additionally, we demonstrate that leveraging redundancy between vision and text can significantly enhance model performance. We hope our study will inform the development of multimodal models to improve spatial intelligence and further close the gap with human intelligence.

Large language models have demonstrated impressive capabilities across various natural language processing tasks, especially in solving mathematical problems. However, large language models are not good at math theorem proving using formal languages like Lean. A significant challenge in this area is the scarcity of training data available in these formal languages. To address this issue, we propose a novel pipeline that iteratively generates and filters synthetic data to translate natural language mathematical problems into Lean 4 statements, and vice versa. Our results indicate that the synthetic data pipeline can provide useful training data and improve the performance of LLMs in translating and understanding complex mathematical problems and proofs. Our final dataset contains about 57K formal-informal question pairs along with searched proof from the math contest forum and 21 new IMO questions. We open-source our code at https://github.com/InternLM/InternLM-Math and our data at https://huggingface.co/datasets/InternLM/Lean-Workbook.

Recent diffusion-based human image animation techniques have demonstrated impressive success in synthesizing videos that faithfully follow a given reference identity and a sequence of desired movement poses. Despite this, there are still two limitations: i) an extra reference model is required to align the identity image with the main video branch, which significantly increases the optimization burden and model parameters; ii) the generated video is usually short in time (e.g., 24 frames), hampering practical applications. To address these shortcomings, we present a UniAnimate framework to enable efficient and long-term human video generation. First, to reduce the optimization difficulty and ensure temporal coherence, we map the reference image along with the posture guidance and noise video into a common feature space by incorporating a unified video diffusion model. Second, we propose a unified noise input that supports random noised input as well as first frame conditioned input, which enhances the ability to generate long-term video. Finally, to further efficiently handle long sequences, we explore an alternative temporal modeling architecture based on state space model to replace the original computation-consuming temporal Transformer. Extensive experimental results indicate that UniAnimate achieves superior synthesis results over existing state-of-the-art counterparts in both quantitative and qualitative evaluations. Notably, UniAnimate can even generate highly consistent one-minute videos by iteratively employing the first frame conditioning strategy. Code and models will be publicly available. Project page: https://unianimate.github.io/., Comment: Project page: https://unianimate.github.io/

Large Language Models (LLMs) struggle with reliably generating highly structured outputs, such as program code, mathematical formulas, or well-formed markup. Constrained decoding approaches mitigate this problem by greedily restricting what tokens an LLM can output at each step to guarantee that the output matches a given constraint. Specifically, in grammar-constrained decoding (GCD), the LLM's output must follow a given grammar. In this paper we demonstrate that GCD techniques (and in general constrained decoding techniques) can distort the LLM's distribution, leading to outputs that are grammatical but appear with likelihoods that are not proportional to the ones given by the LLM, and so ultimately are low-quality. We call the problem of aligning sampling with a grammar constraint, grammar-aligned decoding (GAD), and propose adaptive sampling with approximate expected futures (ASAp), a decoding algorithm that guarantees the output to be grammatical while provably producing outputs that match the conditional probability of the LLM's distribution conditioned on the given grammar constraint. Our algorithm uses prior sample outputs to soundly overapproximate the future grammaticality of different output prefixes. Our evaluation on code generation and structured NLP tasks shows how ASAp often produces outputs with higher likelihood (according to the LLM's distribution) than existing GCD techniques, while still enforcing the desired grammatical constraints.

Owing to the outstanding properties provided by nontrivial band topology, topological phases of matter are considered as a promising platform towards low-dissipation electronics, efficient spin-charge conversion, and topological quantum computation. Achieving ferroelectricity in topological materials enables the non-volatile control of the quantum states, which could greatly facilitate topological electronic research. However, ferroelectricity is generally incompatible with systems featuring metallicity due to the screening effect of free carriers. In this study, we report the observation of memristive switching based on the ferroelectric surface state of a topological semimetal (TaSe4)2I. We find that the surface state of (TaSe4)2I presents out-of-plane ferroelectric polarization due to surface reconstruction. With the combination of ferroelectric surface and charge-density-wave-gapped bulk states, an electric switchable barrier height can be achieved in (TaSe4)2I-metal contact. By employing a multi-terminal grounding design, we manage to construct a prototype ferroelectric memristor based on (TaSe4)2I with on/off ratio up to 10^3, endurance over 10^3 cycles, and good retention characteristics. The origin of the ferroelectric surface state is further investigated by first-principles calculations, which reveals an interplay between ferroelectricity and band topology. The emergence of ferroelectricity in (TaSe4)2I not only demonstrates it as a rare but essential case of ferroelectric topological materials, but also opens new routes towards the implementation of topological materials in functional electronic devices., Comment: 14 pages, 5 figures

Transport probes the motion of quasiparticles in response to external excitations. Apart from the well-known electric and thermoelectric transport, acoustoelectric transport induced by traveling acoustic waves has been rarely explored. Here, by adopting a hybrid nanodevices integrated with piezoelectric substrates, we establish a simple design of acoustoelectric transport with gate tunability. We fabricate dual-gated acoustoelectric devices based on BN-encapsuled graphene on LiNbO3. Longitudinal and transverse acoustoelectric voltages are generated by launching pulsed surface acoustic wave. The gate dependence of zero-field longitudinal acoustoelectric signal presents strikingly similar profiles as that of Hall resistivity, providing a valid approach for extracting carrier density without magnetic field. In magnetic fields, acoustoelectric quantum oscillations appear due to Landau quantization, which are more robust and pronounced than Shubnikov-de Haas oscillations. Our work demonstrates a feasible acoustoelectric setup with gate tunability, which can be extended to the broad scope of various Van der Waals materials., Comment: 16 pages, 5 figures

The evolution of Large Language Models (LLMs) like ChatGPT and GPT-4 has sparked discussions on the advent of Artificial General Intelligence (AGI). However, replicating such advancements in open-source models has been challenging. This paper introduces InternLM2, an open-source LLM that outperforms its predecessors in comprehensive evaluations across 6 dimensions and 30 benchmarks, long-context modeling, and open-ended subjective evaluations through innovative pre-training and optimization techniques. The pre-training process of InternLM2 is meticulously detailed, highlighting the preparation of diverse data types including text, code, and long-context data. InternLM2 efficiently captures long-term dependencies, initially trained on 4k tokens before advancing to 32k tokens in pre-training and fine-tuning stages, exhibiting remarkable performance on the 200k ``Needle-in-a-Haystack" test. InternLM2 is further aligned using Supervised Fine-Tuning (SFT) and a novel Conditional Online Reinforcement Learning from Human Feedback (COOL RLHF) strategy that addresses conflicting human preferences and reward hacking. By releasing InternLM2 models in different training stages and model sizes, we provide the community with insights into the model's evolution.

The math abilities of large language models can represent their abstract reasoning ability. In this paper, we introduce and open-source our math reasoning LLMs InternLM-Math which is continue pre-trained from InternLM2. We unify chain-of-thought reasoning, reward modeling, formal reasoning, data augmentation, and code interpreter in a unified seq2seq format and supervise our model to be a versatile math reasoner, verifier, prover, and augmenter. These abilities can be used to develop the next math LLMs or self-iteration. InternLM-Math obtains open-sourced state-of-the-art performance under the setting of in-context learning, supervised fine-tuning, and code-assisted reasoning in various informal and formal benchmarks including GSM8K, MATH, Hungary math exam, MathBench-ZH, and MiniF2F. Our pre-trained model achieves 30.3 on the MiniF2F test set without fine-tuning. We further explore how to use LEAN to solve math problems and study its performance under the setting of multi-task learning which shows the possibility of using LEAN as a unified platform for solving and proving in math. Our models, codes, and data are released at \url{https://github.com/InternLM/InternLM-Math}.

As a key ingredient in energy harvesting and photodetection, light-to-electricity conversion requires efficient separation of photoexcited electron-hole pairs before recombination. Traditional junction-based mechanisms mainly use build-in electric fields to achieve pair separation and generate photovoltaic effect, which fail to collect photoexcited pairs away from local barrier region. The ability to harvest photovoltaic effect in a homogeneous material upon uniform illumination is appealing, but has only been realized in very few cases such as non-centrosymmetric systems through bulk photovoltaic effect. Here we realize a new type of photovoltaic effect, termed as acousto-drag photovoltaic effect, by travelling surface acoustic waves (t-SAW) in a conventional layered semiconductor MoSe2. Instead of immediately driving the electron-hole pairs to opposite directions after generation, t-SAW induces periodic modulation to electronic bands and drags the photoexcited pairs toward the same travelling direction. The photocurrent can then be extracted by a local barrier, e.g. the metal-semiconductor contact as we used here. By spatially separating the electron-hole generation and extraction processes, the acousto-drag mechanism strongly suppresses charge recombination and yields large nonlocal photoresponse outside the barrier region. We show that when t-SAW is applied, the photoresponse can be enhanced by over two orders of magnitude with exceptionally high external quantum efficiency above 60%. The discovery of acousto-drag photovoltaic effect establishes a new approach towards efficient light-to-electricity conversion without the restriction of crystal symmetry.

Diffusion models have shown remarkable success in a variety of downstream generative tasks, yet remain under-explored in the important and challenging expressive talking head generation. In this work, we propose a DreamTalk framework to fulfill this gap, which employs meticulous design to unlock the potential of diffusion models in generating expressive talking heads. Specifically, DreamTalk consists of three crucial components: a denoising network, a style-aware lip expert, and a style predictor. The diffusion-based denoising network is able to consistently synthesize high-quality audio-driven face motions across diverse expressions. To enhance the expressiveness and accuracy of lip motions, we introduce a style-aware lip expert that can guide lip-sync while being mindful of the speaking styles. To eliminate the need for expression reference video or text, an extra diffusion-based style predictor is utilized to predict the target expression directly from the audio. By this means, DreamTalk can harness powerful diffusion models to generate expressive faces effectively and reduce the reliance on expensive style references. Experimental results demonstrate that DreamTalk is capable of generating photo-realistic talking faces with diverse speaking styles and achieving accurate lip motions, surpassing existing state-of-the-art counterparts., Comment: Project Page: https://dreamtalk-project.github.io

Despite diffusion models having shown powerful abilities to generate photorealistic images, generating videos that are realistic and diverse still remains in its infancy. One of the key reasons is that current methods intertwine spatial content and temporal dynamics together, leading to a notably increased complexity of text-to-video generation (T2V). In this work, we propose HiGen, a diffusion model-based method that improves performance by decoupling the spatial and temporal factors of videos from two perspectives, i.e., structure level and content level. At the structure level, we decompose the T2V task into two steps, including spatial reasoning and temporal reasoning, using a unified denoiser. Specifically, we generate spatially coherent priors using text during spatial reasoning and then generate temporally coherent motions from these priors during temporal reasoning. At the content level, we extract two subtle cues from the content of the input video that can express motion and appearance changes, respectively. These two cues then guide the model's training for generating videos, enabling flexible content variations and enhancing temporal stability. Through the decoupled paradigm, HiGen can effectively reduce the complexity of this task and generate realistic videos with semantics accuracy and motion stability. Extensive experiments demonstrate the superior performance of HiGen over the state-of-the-art T2V methods., Comment: Project page: https://higen-t2v.github.io/

With the advance development in quantum science, constructing a large-scale quantum network has become a hot area of future quantum information technology. Future quantum networks promise to enable many fantastic applications and will unlock fundamentally new technologies in information security and large-scale computation. The future quantum internet is required to connect quantum information processors to achieve unparalleled capabilities in secret communication and enable quantum communication between any two points on Earth. However, the existing quantum networks are basically constructed to realize the communication between the end users in their own networks. How to bridge different independent networks to form a fully-connected quantum internet becomes a pressing challenge for future networks. Here, we demonstrate the quantum fusion of two independent networks for the first time based on multiuser entanglement swapping, to merge two 10-user networks into a larger network with 18 users in quantum correlation layer. By performing the Bell state measurement between two nonneighboring nodes, the users from different networks can establish entanglement and ultimately every pair of the 18 users are able to communicate with each other using the swapped states. Our approach opens attractive opportunities for the establishment of quantum entanglement between remote nodes in different networks, which facilitates versatile quantum information interconnects and has great application in constructing large-scale intercity quantum communication networks.

Video synthesis has recently made remarkable strides benefiting from the rapid development of diffusion models. However, it still encounters challenges in terms of semantic accuracy, clarity and spatio-temporal continuity. They primarily arise from the scarcity of well-aligned text-video data and the complex inherent structure of videos, making it difficult for the model to simultaneously ensure semantic and qualitative excellence. In this report, we propose a cascaded I2VGen-XL approach that enhances model performance by decoupling these two factors and ensures the alignment of the input data by utilizing static images as a form of crucial guidance. I2VGen-XL consists of two stages: i) the base stage guarantees coherent semantics and preserves content from input images by using two hierarchical encoders, and ii) the refinement stage enhances the video's details by incorporating an additional brief text and improves the resolution to 1280$\times$720. To improve the diversity, we collect around 35 million single-shot text-video pairs and 6 billion text-image pairs to optimize the model. By this means, I2VGen-XL can simultaneously enhance the semantic accuracy, continuity of details and clarity of generated videos. Through extensive experiments, we have investigated the underlying principles of I2VGen-XL and compared it with current top methods, which can demonstrate its effectiveness on diverse data. The source code and models will be publicly available at \url{https://i2vgen-xl.github.io}., Comment: Project page: https://i2vgen-xl.github.io