Search

The evolution of machine learning has increasingly prioritized the development of powerful models and more scalable supervision signals. However, the emergence of foundation models presents significant challenges in providing effective supervision signals necessary for further enhancing their capabilities. Consequently, there is an urgent need to explore novel supervision signals and technical approaches. In this paper, we propose verifier engineering, a novel post-training paradigm specifically designed for the era of foundation models. The core of verifier engineering involves leveraging a suite of automated verifiers to perform verification tasks and deliver meaningful feedback to foundation models. We systematically categorize the verifier engineering process into three essential stages: search, verify, and feedback, and provide a comprehensive review of state-of-the-art research developments within each stage. We believe that verifier engineering constitutes a fundamental pathway toward achieving Artificial General Intelligence.

We reveal a quantum coherent state of composite bosonic trions, where paired fermions further pair with bosons, in a one-dimensional Bose-Fermi system featuring onsite boson-fermion attraction and pair hopping of fermions. Extensive density matrix renormalization group (DMRG) calculations show that the quasi-condensation of the composite trions is established by algebraic decaying correlations of trions, gapped single-particle excitations, and suppressed correlations of fermion pairs. Negative binding energy also indicates a strong tendency to form trions. Furthermore, we conduct a phenomenological analysis based on a mean-field variational ansatz to interpret the emergence of composite trions. Additionally, We discuss feasible experimental methods for realizing this novel phase of matter in ultracold atom experiments., Comment: 5 pages,5 figures

Reward Models (RMs) are crucial for aligning language models with human preferences. Currently, the evaluation of RMs depends on measuring accuracy against a validation set of manually annotated preference data. Although this method is straightforward and widely adopted, the relationship between RM accuracy and downstream policy performance remains under-explored. In this work, we conduct experiments in a synthetic setting to investigate how differences in RM measured by accuracy translate into gaps in optimized policy performance. Our findings reveal that while there is a weak positive correlation between accuracy and downstream performance, policies optimized towards RMs with similar accuracy can exhibit quite different performance. Moreover, we discover that the way of measuring accuracy significantly impacts its ability to predict the final policy performance. Through the lens of Regressional Goodhart's effect, we identify the existence of exogenous variables impacting the relationship between RM quality measured by accuracy and policy model capability. This underscores the inadequacy of relying solely on accuracy to reflect their impact on policy optimization.

Distinct from the $\text{SU}(2)$ case, the fermionic systems with $\text{SU}(N)$ symmetry are expected to exhibit novel physics, such as exotic singlet formation. Using the density matrix renormalization group technique, we obtain the ground state phase diagram of the $\text{SU}(3)$ $t$-$J$ chain for density $n<1$. The ground state phase diagram includes the Luttinger liquid, the extended Luther-Emery liquid characterized by a spin gap, and the phase separation state. We quantitatively assess the characteristics of the three phases by measuring spin gap, compressibility, various correlation functions and structure factors. We further study the extended Luther-Emery liquid phase and discover molecular superfluid quasi-long-range order. The mechanism of the molecular superfluid is the combination of three $\text{SU}(3)$ fermions on sites that are not completely connected. Accordingly, we can speculate the behavior of the $\text{SU}(N)$ $t$-$J$ chain model with larger $N$ values, operating within the same filling regime., Comment: 8 pages, 8 figures

Self-critic has become a crucial mechanism for enhancing the reasoning performance of LLMs. However, current approaches mainly involve basic prompts for intuitive instance-level feedback, which resembles System-1 processes and limits the reasoning capabilities. Moreover, there is a lack of in-depth investigations into the relationship between LLM's ability to criticize and its task-solving performance. To address these issues, we propose Critic-CoT, a novel framework that pushes LLMs toward System-2-like critic capability. Through a step-wise CoT reasoning paradigm and the automatic construction of distant-supervision data without human annotation, Critic-CoT enables LLMs to engage in slow, analytic self-critique and refinement, thereby improving their reasoning abilities. Experiments on GSM8K and MATH demonstrate that our enhanced model significantly boosts task-solving performance by filtering out invalid solutions or iterative refinement. Furthermore, we investigate the intrinsic correlation between critique and task-solving abilities within LLMs, discovering that these abilities can mutually reinforce each other rather than conflict., Comment: under review

Vector-matrix-multiplication (VMM) accel-erators have gained a lot of traction, especially due to therise of convolutional neural networks (CNNs) and the desireto compute them on the edge. Besides the classical digitalapproach, analog computing has gone through a renais-sance to push energy efficiency further. A more recent ap-proach is called time-domain (TD) computing. In contrastto analog computing, TD computing permits easy technol-ogy as well as voltage scaling. As it has received limitedresearch attention, it is not yet clear which scenarios aremost suitable to be computed in the TD. In this work, weinvestigate these scenarios, focussing on energy efficiencyconsidering approximative computations that preserve ac-curacy. Both goals are addressed by a novel efficiency met-ric, which is used to find a baseline design. We use SPICEsimulation data which is fed into a python framework toevaluate how performance scales for VMM computation.We see that TD computing offers best energy efficiency forsmall to medium sized arrays. With throughput and sili-con footprint we investigate two additional metrics, givinga holistic comparison., Comment: 8 pages, 12 figures. This paper was accepted at the 25th International Symposium on Quality Electronic Design(ISQED) 2024. DOI: 10.1109/ISQED60706.2024.10528682

We find a novel chiral superfluid (CSF) phase in a one-dimensional Bose-Fermi Hubbard model with significant mass and density imbalance between the two species. In the CSF phase, bosons condensate at non-zero momentum $\pm 2\pi /L$ with chain length $L$. To capture the essential physics of this new phenomenon, we study an alternative simplified model that only features competition between single-fermion hopping and hopping of composite particles composed of a fermion and a boson. This model captures the low energy physics of the Hubbard model and hosts a robust CSF phase. Our unbiased numerical studies show that in the CSF phase, the local superfluid order parameter continuously rotates along the chain, indicating that time-reversal symmetry is spontaneously broken. This symmetry breaking generates an emergent flux in the background, effectively optimizing the system's ground-state energy. We provide a physical understanding at the mean-field level. Furthermore, we have explored the potential realization of this phase in cold-atom experiments., Comment: 6 pages, 5 figures

We extend the single-mode Approximation (SMA) into quantum Monte Carlo simulations to provides an efficient and fast method to obtain the dynamical dispersion of quantum many-body systems. Based on stochastic series expansion (SSE) and its projector algorithms, the SMA + SSE method can simply extract the dispersion of the dynamical dispersion in the long wave-length limit and the upper bound of the dispersion elsewhere, without external calculations and high technique barriers. Meanwhile, numerical analytic continuation methods require the fine data of imaginary time correlations and complex programming. Therefore, our method can approach the excitation dispersion of large systems, e.g., we take the two-dimensional Heisenberg model on a $512 \times 512$ square lattice. We demonstrate the effectiveness and efficiency of our method with high precision via additional examples. We also demonstrate that SMA combined with SSE goes beyond spin-wave theory with numerical results. We further illustrate that SMA is able to extract useful information in strongly correlated systems with competing states.

We use the density matrix renormalization group (DMRG) method to study the SU(4) $t$-$J$ chain. We find that, in addition to the conventional repulsive Luttinger liquid phase and phase separation, there are two phases in the attractive Luttinger liquid region dependent on whether the flavor gap is opened or not. The first with the flavor gap is the molecular superfluid phase (the SU(4) singlet instability) which is well-known in the attractive SU(4) Hubbard model ($U<0$). The second without the flavor gap is the superconducting phase (the six-component pairing instability). Furthermore, the molecular superfluid instability cannot coexist with the superconducting instability. This is general in SU($N$) models with $N>2$ and is well demonstrated by the theoretical analysis based on the phenomenological bosonization results., Comment: 11 pages, 6 figures

Named entity recognition in real-world applications suffers from the diversity of entity types, the emergence of new entity types, and the lack of high-quality annotations. To address the above problems, this paper proposes an in-context learning-based NER approach, which can effectively inject in-context NER ability into PLMs and recognize entities of novel types on-the-fly using only a few demonstrative instances. Specifically, we model PLMs as a meta-function $\mathcal{ \lambda_ {\text{instruction, demonstrations, text}}. M}$, and a new entity extractor can be implicitly constructed by applying new instruction and demonstrations to PLMs, i.e., $\mathcal{ (\lambda . M) }$(instruction, demonstrations) $\to$ $\mathcal{F}$ where $\mathcal{F}$ will be a new entity extractor, i.e., $\mathcal{F}$: text $\to$ entities. To inject the above in-context NER ability into PLMs, we propose a meta-function pre-training algorithm, which pre-trains PLMs by comparing the (instruction, demonstration)-initialized extractor with a surrogate golden extractor. Experimental results on 4 few-shot NER datasets show that our method can effectively inject in-context NER ability into PLMs and significantly outperforms the PLMs+fine-tuning counterparts., Comment: Accepted to ACL 2023 Main Conference

The challenge of information extraction (IE) lies in the diversity of label schemas and the heterogeneity of structures. Traditional methods require task-specific model design and rely heavily on expensive supervision, making them difficult to generalize to new schemas. In this paper, we decouple IE into two basic abilities, structuring and conceptualizing, which are shared by different tasks and schemas. Based on this paradigm, we propose to universally model various IE tasks with Unified Semantic Matching (USM) framework, which introduces three unified token linking operations to model the abilities of structuring and conceptualizing. In this way, USM can jointly encode schema and input text, uniformly extract substructures in parallel, and controllably decode target structures on demand. Empirical evaluation on 4 IE tasks shows that the proposed method achieves state-of-the-art performance under the supervised experiments and shows strong generalization ability in zero/few-shot transfer settings., Comment: accepted by AAAI2023

We examine the effect of boson-fermion interaction in a one-dimensional Bose-Fermi mixture by using the density matrix renormalization group method. We show that the boson superfluidity is enhanced by fermions for a weak boson-fermion coupling at an approximate integer boson filling factor (e.g., $0.935\le \rho_b \le 1.0$), and this enhancement is produced both in a fermion metallic state and in a fermion insulating state. A metal-insulator phase transition of fermions induced by boson-fermion interaction is observed even though there is no fermion-fermion interaction in the parent Hamiltonian. Furthermore, we find that the boson superfluid order and density wave order can coexist in a deep fermion Mott region. All these features could be measured in future experiments and open up the possibility of detecting the new physical effect in the Bose-Fermi mixture., Comment: 7 pages, 5(3+2) figures

We study the two-dimensional Bose-Fermi mixture on square lattice at finite temperature by using the determinant quantum Monte Carlo method within the weakly interacting regime. Here we consider the attractive Bose-Hubbard model and free spinless fermions. In the absence of bosonfermion interactions, we obtain the boundary of the collapsed state of the attractive bosons. In the presence of boson-fermion interactions, an effective p-wave interaction between fermions will be induced as far as the bosons are in a superfluid state. Moreover, we find the emergence of the composite fermion pairs at low temperatures.

Since constraints hinder the application of numerical algorithms, phase diagrams of quantum dimer models are still controversial, even on the square lattice. The core controversy is whether the mixed state exists. In this article, we give strong evidences to solve this dispute. With our newly developed sweeping cluster method, we studied the phase diagram of large parameter space by introducing the definition of pair correlation function and other supporting evidence to distinguish the mixed phase from the columnar phase with high precision. In particular, we find that the ground state corresponds to the mixed phase for a vast parameter region.

Spoken language understanding (SLU) acts as a critical component in goal-oriented dialog systems. It typically involves identifying the speakers intent and extracting semantic slots from user utterances, which are known as intent detection (ID) and slot filling (SF). SLU problem has been intensively investigated in recent years. However, these methods just constrain SF results grammatically, solve ID and SF independently, or do not fully utilize the mutual impact of the two tasks. This paper proposes a multi-head self-attention joint model with a conditional random field (CRF) layer and a prior mask. The experiments show the effectiveness of our model, as compared with state-of-the-art models. Meanwhile, online education in China has made great progress in the last few years. But there are few intelligent educational dialog applications for students to learn foreign languages. Hence, we design an intelligent dialog robot equipped with different scenario settings to help students learn communication skills.

Quantum spin systems with strong geometric restrictions give rise to rich quantum phases such as valence bond solids and spin liquid states. However, the geometric restrictions often hamper the application of sophisticated numerical approaches. Based on the stochastic series expansion method, we develop an efficient and exact quantum Monte Carlo "sweeping cluster" algorithm which automatically satisfies the geometrical restrictions. Here we use the quantum dimer model as a benchmark to demonstrate the reliability and power of this algorithm. Comparing to existing numerical methods, we can obtain higher accuracy results for a wider parameter region and much more substantial system sizes.

This research uses Twitter, as a social media device, to track communications related to the 2015 U.S. foodborne illness outbreak linked to Salmonella in imported cucumbers from Mexico. The relevant Twitter data are analyzed in light of the timeline of the official announcements made by the Centers for Disease Control and Prevention (CDC). The largest number of registered tweets is associated with the period immediately following the CDC initial announcement and the official release of the first recall of cucumbers.

Quantum dissipation arises when a large system can be split in a quantum system and an environment where the energy of the former flows to. Understanding the effect of dissipation on quantum many-body systems is of particular importance due to its potential relations with quantum information processing. We propose a conceptually simple approach to introduce the dissipation into interacting quantum systems in a thermodynamical context, in which every site of a 1d lattice is coupled off-diagonally to its own bath. The interplay between quantum dissipation and interactions gives rise to counterintuitive interpretations such as a compressible zero-temperature state with spontaneous discrete symmetry breaking and a thermal phase transition in a one-dimensional dissipative quantum many-body system as revealed by Quantum Monte Carlo path integral simulations.