Searcharxiv⌕ Search

arXiv subjects

Yang Gao

Publications and source records attributed to Yang Gao.

At least 523 records · Page 29Linked to original sources
arXiv2023

Policy Contrastive Imitation Learning

Adversarial imitation learning (AIL) is a popular method that has recently achieved much success. However, the performance of AIL is still unsatisfactory on the more challenging tasks. We find that one of the major reasons is due to the low quality of AIL discriminator representation. Since the AIL discriminator is trained via binary classification that does not necessarily discriminate the policy from the expert in a meaningful way, the resulting reward might not be meaningful either. We propose a new method called Policy Contrastive Imitation Learning (PCIL) to resolve this issue. PCIL learns a contrastive representation space by anchoring on different policies and generates a smooth cosine-similarity-based reward. Our proposed representation learning objective can be viewed as a stronger version of the AIL objective and provide a more meaningful comparison between the agent and the policy. From a theoretical perspective, we show the validity of our method using the apprenticeship learning framework. Furthermore, our empirical evaluation on the DeepMind Control suite demonstrates that PCIL can achieve state-of-the-art performance. Finally, qualitative results suggest that PCIL builds a smoother and more meaningful representation space for imitation learning.

cs.LG↗
arXiv2023

Measurement of $e^+e^-\to pK^-\barΛ+c.c.$ cross sections between 4.009 GeV and 4.951 GeV

Using $e^+e^-$ collision datasets corresponding to total integrated luminosity of 21.7 fb$^{-1}$ collected with the BESIII detector at the BEPCII collider at center-of-mass energies ranging from 4.009 GeV to 4.951 GeV, the energy-dependent cross sections of $e^+e^-\to pK^-\barΛ+c.c.$ are measured for the first time. By fitting these energy-dependent cross sections, we search for the excited $ψ$ states $ψ(4160)$ and $ψ(4415)$, and the vector charmonium-like states $ψ(4230)$, $ψ(4360)$, and $ψ(4660)$. No evidence for these is observed and the upper limits on the branching fractions of these states decaying into $pK^-\bar Λ+c.c.$ are set at the 90\% confidence level.

hep-ex↗
arXiv2023

Search for the semi-leptonic decays $Λ_c^+ \to Λπ^+ π^- e^+ ν_e$ and $Λ_c^+ \to p K_S^0 π^- e^+ ν_e$

We search for the semi-leptonic decays $Λ_c^+ \to Λπ^+ π^- e^+ ν_e$ and $Λ_c^+ \to p K_S^0 π^- e^+ ν_e$ in a sample of 4.5 $\mathrm{fb}^{-1}$ of $e^{+}e^{-}$ annihilation data collected in the center-of-mass energy region between 4.600 GeV and 4.699 GeV by the BESIII detector at the BEPCII. No significant signals are observed, and the upper limits on the decay branching fractions are set to be $\mathcal{B}(Λ_c^+ \to Λπ^+ π^- e^+ ν_e)<3.9\times10^{-4}$ and $\mathcal{B}(Λ_c^+ \to p K_S^0 π^- e^+ ν_e)<3.3\times10^{-4}$ at the 90% confidence level, respectively.

hep-ex↗
arXiv2023

For Pre-Trained Vision Models in Motor Control, Not All Policy Learning Methods are Created Equal

In recent years, increasing attention has been directed to leveraging pre-trained vision models for motor control. While existing works mainly emphasize the importance of this pre-training phase, the arguably equally important role played by downstream policy learning during control-specific fine-tuning is often neglected. It thus remains unclear if pre-trained vision models are consistent in their effectiveness under different control policies. To bridge this gap in understanding, we conduct a comprehensive study on 14 pre-trained vision models using 3 distinct classes of policy learning methods, including reinforcement learning (RL), imitation learning through behavior cloning (BC), and imitation learning with a visual reward function (VRF). Our study yields a series of intriguing results, including the discovery that the effectiveness of pre-training is highly dependent on the choice of the downstream policy learning algorithm. We show that conventionally accepted evaluation based on RL methods is highly variable and therefore unreliable, and further advocate for using more robust methods like VRF and BC. To facilitate more universal evaluations of pre-trained models and their policy learning methods in the future, we also release a benchmark of 21 tasks across 3 different environments alongside our work.

cs.CV↗
arXiv2023

Measurements of Differential Cross Sections of Inclusive $π^0$ and $K^0_S$ Production in $e^{+}e^{-}$ Annihilation at Energies from 2.2324 to 3.6710 GeV

Based on electron positron collision data collected with the BESIII detector operating at the BEPCII storage rings, the differential cross sections of inclusive $π^0$ and $K^0_S$ production as a function of hadron momentum, normalized by the total cross section of the $e^{+}e^{-} \to$ hadrons process, are measured at six center-of-mass energies from 2.2324 to 3.6710 GeV. Our results with a relative hadron energy coverage from 0.1 to 0.9 significantly deviate from several theoretical calculations based on existing fragmentation functions, especially at lower energies.

hep-ex↗
arXiv2023

Study of $Λ_c^+\rightarrow Λμ^+ν_μ$ and Test of Lepton Flavor Universality with $Λ_c^+\rightarrow Λ\ell^+ν_{\ell}$ Decays

The measurement of the Cabibbo-favored semileptonic decay $Λ_c^+\rightarrow Λμ^+ν_μ$ is reported using $4.5~\mathrm{fb}^{-1}$ of $e^+e^-$ annihilation data collected at center-of-mass energies ranging from 4.600~GeV to 4.699~GeV. The branching fraction of the decay is measured to be $\mathcal{B}(Λ_c^+\rightarrow Λμ^+ν_μ)=(3.48\pm0.14_{\rm stat.}\pm0.10_{\rm syst.})\%$, three times more precise than the prior world average result. Tests of lepton flavor universality using $Λ_c^+\rightarrow Λ\ell^+ν_{\ell}$ ($\ell=e, μ$) decays are reported for the first time, based on measurements of the differential decay rates and the forward-backward asymmetries in separate four-momentum transfer regions. The results are compatible with Standard Model predictions. Furthermore, we improve the determination of the form-factor parameters in $Λ_c^+\rightarrow Λ\ell^+ν_{\ell}$ decays, which provide stringent tests and calibration for lattice quantum chromodynamics (LQCD) calculations.

hep-ex↗
arXiv2023

Berry-Curvature Engineering for Nonreciprocal Directional Dichroism in Two-Dimensional Antiferromagnets

In two-dimensional antiferromagnets, we identify the mixed Berry curvature as the geometrical origin of the nonreciprocal directional dichroism (NDD), which refers to the difference in light absorption with the propagation direction flipped. Such a Berry curvature is strongly tied to the uniaxial strain in accordance with the symmetry constraint, leading to a highly tunable NDD, whose sign and magnitude can be manipulated via the strain direction. As a concrete example, we demonstrate such a phenomenon in a lattice model of MnBi2Te4. The coupling between the mixed Berry curvature and strain also suggests the magnetic quadrupole of the Bloch wave packet as the macroscopic order parameter probed by the NDD in two dimensions, distinct from the multiferroic order P times M or the spin toroidal and quadrupole order within a unit cell in previous studies. Our work paves the way of the Berry-curvature engineering for optical nonreciprocity in two-dimensional antiferromagnets.

cond-mat.mes-hall↗
arXiv2023

First study of reaction $Ξ^{0}n\rightarrowΞ^{-}p$ using $Ξ^0$-nucleus scattering at an electron-positron collider

Using $(1.0087\pm0.0044)\times10^{10}$ $J/ψ$ events collected with the BESIII detector at the BEPCII storage ring, the process $Ξ^{0}n\rightarrowΞ^{-}p$ is studied, where the $Ξ^0$ baryon is produced in the process $J/ψ\rightarrowΞ^0\barΞ^0$ and the neutron is a component of the $^9\rm{Be}$, $^{12}\rm{C}$ and $^{197}\rm{Au}$ nuclei in the beam pipe. A clear signal is observed with a statistical significance of $7.1σ$. The cross section of the reaction $Ξ^0+{^9\rm{Be}}\rightarrowΞ^-+p+{^8\rm{Be}}$ is determined to be $σ(Ξ^0+{^9\rm{Be}}\rightarrowΞ^-+p+{^8\rm{Be}})=(22.1\pm5.3_{\rm{stat}}\pm4.5_{\rm{sys}})$ mb at the $Ξ^0$ momentum of $0.818$ GeV/$c$, where the first uncertainty is statistical and the second is systematic. No significant $H$-dibaryon signal is observed in the $Ξ^-p$ final state. This is the first study of hyperon-nucleon interactions in electron-positron collisions and opens up a new direction for such research.

hep-ex↗
arXiv2023

Determination of spin and parity of $D^{*}_{(s)}$ mesons

The spin and parity of the charmed mesons $D_{s}^{*+}$, $D^{*0}$ and $D^{*+}$ are determined for the first time to be $J^P=1^{-}$ with significances greater than 10$σ$ over other hypotheses of $2^{+}$ and $3^{-}$, using an $e^+e^-$ collision data sample with an integrated luminosity of 3.19 fb$^{-1}$ collected by the BESIII detector at a center-of-mass energy of 4.178 GeV. Different spin-parity hypotheses for $D_{s}^{*+}$, $D^{*0}$, and $D^{*+}$ mesons are tested via a helicity amplitude analysis of the processes $e^+e^-\to D^{*+}_{s}D^{-}_{s}$, $D^{*0}D^{0}$ and $D^{*+}D^{-}$, with $D^{*+}_{s}\to D^{+}_{s} γ$, $D^{*0}\to D^{0}π^{0}$, and $D^{*+}\to D^{+}π^{0}$. The results confirm the quark model predictions.

hep-ex↗
arXiv2023

Precision Measurement of the Decay $Σ^{+} \rightarrow pγ$ in the Process $J/ψ\rightarrow Σ^{+}\barΣ^{-}$

Using $(10\:087\pm44)\times10^{6}$ $J/ψ$ events collected with the BESIII detector, the radiative hyperon decay $Σ\rightarrow pγ$ is studied at an electron-positron collider experiment for the first time. The absolute branching fraction is measured to be $\left(0.996 \pm 0.021_{\rm stat.}\pm 0.018_{\rm syst.}\right)\times 10^{-3}$, which is lower than its world average value by $4.2$ standard deviations. Its decay asymmetry parameter is determined to be $-0.652 \pm 0.056_{\rm stat.}\pm 0.020_{\rm syst.}$. The branching fraction and decay asymmetry parameter are the most precise to date, and the accuracies are improved by $78\%$ and $34\%$, respectively.

hep-ex↗
arXiv2023

Precision measurement of the matrix elements for $η\toπ^+π^-π^0$ and $η\toπ^0π^0π^0$ decays

A precision measurement of the matrix elements for $η\toπ^+π^-π^0$ and $η\toπ^0π^0π^0$ decays is performed using a sample of $(10087\pm44)\times10^6$ $J/ψ$ decays collected with the BESIII detector. The decay $J/ψ\to γη$ is used to select clean samples of 631,686 $η\toπ^+π^-π^0$ decays and 272,322 $η\toπ^0π^0π^0$ decays. The matrix elements for both channels are in reasonable agreement with previous measurements. The non-zero $gX^2Y$ term for the decay mode $η\toπ^+π^-π^0$ is confirmed, as reported by the KLOE Collaboration, while the other higher-order terms are found to be insignificant. Dalitz plot asymmetries in the $η\toπ^+π^-π^0$ decay are also explored and are found to be consistent with charge conjugation invariance. In addition, a cusp effect is investigated in the $η\toπ^0π^0π^0$ decay, and no obvious structure around the $π^+π^-$ mass threshold is observed.

hep-ex↗
arXiv2023

Search for baryon and lepton number violating decays of $Ξ^{0}$ hyperons

Using $(1.0087\pm0.0044)\times10^{10}$ $J/ψ$ events collected by the BESIII detector at the BEPCII collider, we report the first search for the baryon and lepton number violating decays $Ξ^{0} \rightarrow K^{-} e^{+}$ with $Δ(B-L)=0$ and $Ξ^{0} \rightarrow K^{+} e^{-}$ with $|Δ(B-L)|=2$, where $B$ ($L$) is the baryon (lepton) number. While no signal is observed, the upper limits on the branching fractions of these two decays are set to $\mathcal B(Ξ^{0} \rightarrow K^{-} e^{+})<3.6\times10^{-6}$ and $\mathcal B(Ξ^{0} \rightarrow K^{+} e^{-})<1.9\times10^{-6}$ at the 90\% confidence level, respectively. These results offer a direct probe of baryon number violating interactions involving a strange quark.

hep-ex↗
arXiv2023

Determination of the $C\!P$-even fraction of $D^0\rightarrow K_S^0π^+π^-π^0$

Quantum-correlated $D\bar{D}$ pairs collected by the BESIII experiment at the $ψ(3770)$ resonance, corresponding to an integrated luminosity of 2.93 fb$^{-1}$, are used to study the $D^0 \rightarrow K^{0}_Sπ^{+} π^{-} π^{0}$ decay mode. The $C\!P$-even fraction of $D^0 \rightarrow K^{0}_Sπ^{+} π^{-} π^{0}$ decays is determined to be $0.235\pm 0.010\pm 0.002$, where the first uncertainty is statistical and the second is systematic.

hep-ex↗
arXiv2023

MPMNet: A Data-Driven MPM Framework for Dynamic Fluid-Solid Interaction

High-accuracy, high-efficiency physics-based fluid-solid interaction is essential for reality modeling and computer animation in online games or real-time Virtual Reality (VR) systems. However, the large-scale simulation of incompressible fluid and its interaction with the surrounding solid environment is either time-consuming or suffering from the reduced time/space resolution due to the complicated iterative nature pertinent to numerical computations of involved Partial Differential Equations (PDEs). In recent years, we have witnessed significant growth in exploring a different, alternative data-driven approach to addressing some of the existing technical challenges in conventional model-centric graphics and animation methods. This paper showcases some of our exploratory efforts in this direction. One technical concern of our research is to address the central key challenge of how to best construct the numerical solver effectively and how to best integrate spatiotemporal/dimensional neural networks with the available MPM's pressure solvers.

cs.GR↗
arXiv2023

Test of $C\!P$ Symmetry in Hyperon to Neutron Decays

The quantum entangled $J/ψ\to Σ^{+}\barΣ^{-}$ pairs from $(1.0087\pm0.0044)\times10^{10}$ $J/ψ$ events taken by the BESIII detector are used to study the non-leptonic two-body weak decays $Σ^{+} \to n π^{+}$ and $\barΣ^{-} \to \bar{n} π^{-}$. The $C\!P$-odd weak decay parameters of the decays $Σ^{+} \to n π^{+}$ ($α_{+}$) and $\barΣ^{-} \to \bar{n} π^{-}$ ($\barα_{-}$) are determined to be $-0.0565\pm0.0047_{\rm stat}\pm0.0022_{\rm syst}$ and $0.0481\pm0.0031_{\rm stat}\pm0.0019_{\rm syst}$, respectively. The decay parameter $\barα_{-}$ is measured for the first time, and the accuracy of $α_{+}$ is improved by a factor of four compared to the previous results. The simultaneously determined decay parameters allow the first precision $C\!P$ symmetry test for any hyperon decay with a neutron in the final state with the measurement of $A_{C\!P}=(α_{+}+\barα_{-})/(α_{+}-\barα_{-})=-0.080\pm0.052_{\rm stat}\pm0.028_{\rm syst}$. Assuming $C\!P$ conservation, the average decay parameter is determined as $\left< α_{+}\right>=(α_{+}- \barα_{-})/2 = -0.0506\pm0.0026_{\rm stat}\pm0.0019_{\rm syst}$, while the ratios $α_{+}/α_{0}$ and $\barα_{-}/\barα_{0}$ are $-0.0490\pm0.0032_{\rm stat}\pm0.0021_{\rm syst}$ and $-0.0571\pm0.0053_{\rm stat}\pm0.0032_{\rm syst}$, where $α_{0}$ and $\barα_{0}$ are the decay parameters of the decays $Σ^{+} \to p π^{0}$ and $\barΣ^{-} \to \bar{p} π^{0}$, respectively.

hep-ex↗
arXiv2023

Shades of meaning: Uncovering the geometry of ambiguous word representations through contextualised language models

Lexical ambiguity presents a profound and enduring challenge to the language sciences. Researchers for decades have grappled with the problem of how language users learn, represent and process words with more than one meaning. Our work offers new insight into psychological understanding of lexical ambiguity through a series of simulations that capitalise on recent advances in contextual language models. These models have no grounded understanding of the meanings of words at all; they simply learn to predict words based on the surrounding context provided by other words. Yet, our analyses show that their representations capture fine-grained meaningful distinctions between unambiguous, homonymous, and polysemous words that align with lexicographic classifications and psychological theorising. These findings provide quantitative support for modern psychological conceptualisations of lexical ambiguity and raise new challenges for understanding of the way that contextual information shapes the meanings of words across different timescales.

cs.CL↗
arXiv2023

Programmatically Grounded, Compositionally Generalizable Robotic Manipulation

Robots operating in the real world require both rich manipulation skills as well as the ability to semantically reason about when to apply those skills. Towards this goal, recent works have integrated semantic representations from large-scale pretrained vision-language (VL) models into manipulation models, imparting them with more general reasoning capabilities. However, we show that the conventional pretraining-finetuning pipeline for integrating such representations entangles the learning of domain-specific action information and domain-general visual information, leading to less data-efficient training and poor generalization to unseen objects and tasks. To this end, we propose ProgramPort, a modular approach to better leverage pretrained VL models by exploiting the syntactic and semantic structures of language instructions. Our framework uses a semantic parser to recover an executable program, composed of functional modules grounded on vision and action across different modalities. Each functional module is realized as a combination of deterministic computation and learnable neural networks. Program execution produces parameters to general manipulation primitives for a robotic end-effector. The entire modular network can be trained with end-to-end imitation learning objectives. Experiments show that our model successfully disentangles action and perception, translating to improved zero-shot and compositional generalization in a variety of manipulation behaviors. Project webpage at: \url{https://progport.github.io}.

cs.AI↗
arXiv2023

Measurement of $e^+ e^- \rightarrow ϕη^{\prime}$ cross sections at center-of-mass energies between 3.508 and 4.600 GeV

We present a measurement of the dressed cross sections for $e^+ e^- \rightarrow ϕη^{\prime}$ at different center-of-mass energies between 3.508 and 4.600 GeV based on 15.1 fb$^{-1}$ of $e^+ e^-$ annihilation data collected with the BESIII detector operating at the BEPCII collider. In addition, a search for the decay $Y(4230) \to ϕη^{\prime}$ is performed. No clear signal is observed and the corresponding upper limit is provided.

hep-ex↗