Search

In-context learning (ICL) with pre-trained language models (PTLMs) has shown great success in code generation. ICL does not require training. PTLMs take as the input a prompt consisting of a few requirement-code examples and a new requirement, and output a new program. However, existing studies simply reuse ICL techniques for natural language generation and ignore unique features of code generation. We refer to these studies as standard ICL. Inspired by observations of the human coding process, we propose a novel ICL approach for code generation named AceCoder. Compared to standard ICL, AceCoder has two novelties. (1) Example retrieval. It retrieves similar programs as examples and learns programming skills (e.g., algorithms, APIs) from them. (2) Guided Code Generation. It encourages PTLMs to output an intermediate preliminary (e.g., test cases, APIs) before generating programs. The preliminary can help PTLMs understand requirements and guide the next code generation. We apply AceCoder to six PTLMs (e.g., Codex) and evaluate it on three public benchmarks using the Pass@k. Results show that AceCoder can significantly improve the performance of PTLMs on code generation. (1) In terms of Pass@1, AceCoder outperforms standard ICL by up to 79.7% and fine-tuned models by up to 171%. (2) AceCoder is effective in PTLMs with different sizes (e.g., 1B to 175B) and different languages (e.g., Python, Java, and JavaScript). (3) We investigate multiple choices of the intermediate preliminary. (4) We manually evaluate generated programs in three aspects and prove the superiority of AceCoder. (5) Finally, we discuss some insights about ICL for practitioners.

The state of health (SOH) is essential for the safety and reliability of lithium-ion batteries in different applications. Moreover, the inconsistency of battery cells in battery packs is inevitable. To determine the SOH difference is even more challenging without the calculation of the SOH of each battery cell. To solve the problem, this article proposes a novel relative SOH (rSOH) estimation method based on wavelet analysis, and the SOH differences for series-connected lithium-ion battery cells in the battery pack can be obtained. It should be noted that no battery model is needed in this method. The more accessible battery voltage of each battery cell is sufficient to obtain the SOH differences, without SOH estimation for each battery cell. With the wavelet analysis, more detailed information of the battery cells is obtained. The rSOH is then introduced as the index of the SOH difference, and the estimation procedure is further analyzed. Finally, experimental validations are conducted by different aged battery cells. The different state of charge conditions and different driving cycles are applied to show the robustness of the proposed method. The experimental results show that the rSOH estimation error is limited within 5%, equivalently with less than 1% SOH estimation error, thus proving the effectiveness of the proposed rSOH estimation method.

For the current stateful network protocol fuzzing technology, the auxiliary type message repeated interaction affects the test efficiency and ensures the validity of the test case by inputting the corresponding test case according to the state of the protocol entity, so that the message abnormality input sequence test cannot be performed. In this paper, a stateful network protocol fuzzing method is designed based on Q-learning algorithm. The auxiliary message of the boot state is not required, and the message abnormality input sequence test can be performed under the premise of ensuring the validity of the test case. Experimental results show that this fuzzing method can significantly improve test efficiency and vulnerability mining capabilities.

High-value bio-oil and gaseous products can be produced by pyrolysis of corn stover. This work presents a systematic study on pyrolysis characteristics of corn stover using an innovative infrared heated pyrolysis reactor. Firstly, the analytical pyrolysis was performed in a thermogravimetric (TG) analyzer at 10, 20, 30 and 40 °C/min to investigate the thermal degradation behaviors, and the pyrolysis kinetics and thermodynamics parameters were calculated. The activation energies calculated by FWO and KAS method were 19.61 ∼ 38.33 kJ/mol and 11.40 ∼ 29.51 kJ/mol respectively at conversion fractions (α) ranging from 0.2 to 0.8, and the one-dimensional diffusion (D1) has the highest fitting degree. The infrared heated technology effectively suppressed the secondary reaction of primary volatiles, and it presented a higher bio-oil yield than electrically heated technology, and as a result, the char and gas yields had a lower value. In both reactors, the yield of bio-oil increases at first and then decreases with the increase in temperature, and the bio-oil yield of the infrared heated furnace and electrically heated furnace reaches the maximum value at 550 ℃ and 500 ℃, which are 34.34 wt.% and 26.40 wt.%, respectively. The acids content of bio-oil decreased gradually with the temperature increased in the infrared heated reactor. The yield of phenolic compounds in bio-oil was higher in the former than in the latter, at 19.36% and 14.86%, respectively. The infrared heated reactor has superior performance in producing high quality bio-oils compared to conventional electric heating methods and may be considered as a promising biomass conversion technique.

Vegetation is widely used for reinforcing slopes, and thus controlling shallow landslides, because plant root systems can increase soil shear strength by anchoring soil layers and modifying hydrological characteristics. However, limited information is available regarding the vertical variation of soil reinforcement by roots at the slope scale, and how slope stability is influenced, especially in plantation areas. In this paper we examine the biomechanical effects on slope stability of the roots of Populus tomentosa, Robinia pseudoacacia and Olea europaea in a geohazard-prone region in China. Root density, variation in root area ratio (RAR) with depth, root tensile strength and root peak tensile force are measured, along with calculating the contribution of root cohesion to the factor of safety. The Wu-Waldron model and fiber bundle model are used to estimate static stress and dynamic stress adherence of plant roots to soil during debris flows and landslides, and a two-dimensional finite element model is used to calculate the increase in the factor of safety caused by additional root cohesion. The results show that RAR is higher for P. tomentosa than for R. pseudoacacia and O. europaea. The vertical soil profile of R. pseudoacacia roots has the highest soil water content (7.06%). The additional cohesion provided by R. pseudoacacia roots (10.35–15.12 kPa) is greater than that provided by P. tomentosa roots (6.19–10.26 kPa) and O. europaea roots (2.31–5.06 kPa). The stable area is also larger on a slope planted with R. pseudoacacia compared with the other two species. Young growth roots (8 years) are shown to affect the distribution of soil water and are limited by soil water in the semi-humid region. R. pseudoacacia roots provide significant additional cohesion by both static and dynamic stress for slope stability. This study aims to increase current understanding of the biomechanical properties of plantation species roots and their efficacy in soil reinforcement.

It is of great important to obtain the states of the batteries in the storage system in Electric Vehicle, wind, solar power and so on. Among these states, the State of Health (SOH) is particularly important and drawing more and more attention. This paper proposes the wavelet based relative SOH (rSOH) estimation method, especially for the battery storage systems where battery cells are connected in serials. With the rSOH, the fading differences in the battery string can be obtained, which is very useful for battery balancing. By taking advantage of the wavelet decomposition, the battery voltage is decomposed into several layers and the characteristics of the relationship between the decomposed signals and the rSOH is analyzed. The wavelet based rSOH estimation method is then proposed and analysed. The experimental validation is carried out and the rSOH estimation results is proved to be effective.

In this paper, Ensemble-Based Graph Model and Dense Reconstruction Error for Infrared Target Detection algorithm is proposed. Firstly, the infrared image is constructed as a closed-loop by using the super-pixel segmentation. Then the saliency map of the target area and the corresponding part of the background are extracted respectively by using the graph model-based manifold ranking algorithm and the dense reconstruction error. And the fused result of the two saliency maps is insensitive to the background clutter interference, which can clearly locate the target area. The experimental results show that the proposed algorithm can suppress the background clutter interference and maintain the integrity of the target edge.

This study investigates why newly listed firms become M&A targets shortly after their initial public offering (IPOs) from the perspective of product innovation. We find strong empirical evidence that IPOs with less established trademarks increase the likelihood of becoming IPO targets. We also find that the negative relation between established trademarks and the likelihood of becoming IPO targets is more pronounced in highly competitive industries and is primarily driven by the M&A supply side. IPOs with more established trademarks can fend themselves against the product market race as independent firms. They can meanwhile realize superior post-IPO financial as well as innovation performance. To acquire such firms, acquirers need to offer substantially higher takeover premiums. However, some empirical evidence suggests that less product innovation-intensive IPOs tend to deliberately seek potential acquirers to support their product market competing position and therefore are more likely to initiate an M&A deal shortly after going public.

A robust d3-methylating agent DMTT has been developed, and its selectivity in d3-methylation of complex molecules evaluated., Methods to incorporate deuterium atoms into organic molecules are valuable for the pharmaceutical industry. The introduction of deuterium atoms by a synthetic method enables the direct tracing of the drug molecule without substantially altering its structure or function. The methyl group is one of the most commonly occurring carbon fragments in biologically active molecules. Here, a biomimetic design reagent, 5-(methyl-d3)-5H-dibenzo[b,d]thiophen-5-ium trifluoromethane sulfonate (DMTT), as an analog of S-adenosylmethionine (SAM), has been developed for the selective d3-methylation of complex molecules bearing several possible reactive sites with excellent selectivity and high-level deuterium incorporation. A series of d3-methylated organic molecules and deuterated pharmaceuticals were synthesized under the mild system with excellent functional group compatibility.

The adaptive fading extended Kalman filter (AFEKF) is proposed for the accuracy and convergent speed problem of state of charge (SOC) estimation. This filter algorithm combines the adaptive extended Kalman filter and fading extended Kalman filter, which solves the problem of uncertainty of system noise and over-reliance on old data. The equivalent circuit model is utilized and the variable forgetting factor recursive least square(VFFRLS) is adopted to identify the model parameters. Experimental test platform is established to validate the method. Results show that the method is able to effectively improve the convergent speed and accuracy of SOC estimation, and the SOC error is less than 2%.

Correct non-standard protocol format recognition is the foundation of protocol security analysis, which is an important part of industrial control system(ICS) information security content. Due to current situation of ICS and protocol features of structure determination, transmission repeat and semantic limited, a method based on net-trace is proposed. The formats of protocol are gotten by single message processing for a preliminary clustering, packet processing for sequence alignment, key fields to infer semantics. Verification results show that the method can reverse recognition non-standard ICS protocol format.

The accumulation and dynamic stock of soil organic carbon (SOC) in agricultural land are highly sensitive to changes in climate. However, there is uncertainty about how the global SOC cycle will respond to future climate change. In order to assess changes in SOC storage, basing on the historical weather data and field observations (1951–2018), the study used the DeNitrification-DeComposition (DNDC) model to simulate future SOC storage in an arid oasis under three representational concentration paths (RCPs): RCP2.6, low-end path; RCP4.5, middle stable path; and RCP8.5, high-end path) for the period from 2018 to 2086. Farmland soil in the Yanqi Basin in northern China was divided into grid units using ArcGIS, and then kriging interpolation was used to extract the mean value of the 200 grid units to calculate regional changes in SOC storage. The model indicated that more than 92.75 % of the measured results accurately simulated the SOC storage in Yanqi Basin cropland. The measured SOC storage in the Yanqi Basin was 1.32 Tg C in 2018. The range of SOC density was 653.27–5527.04 kg C/hm2 under the historical climate scenario, and 1863.99–4581.16 kg C/hm2, 2372.74–6910.15 kg C/hm2, and 623.656–5431.95 kg C/hm2 under the RCP 2.6, RCP 4.5, and RCP 8.5 scenarios, respectively. The simulation results show that, compared with the SOC stock in 2018, the SOC stocks in the oasis area in the future will show a rapid decline under the historical climate scenario (7.6 % decrease) as well as in the RCP 2.6 (3% decrease), RCP 4.5 (9.17 % decrease), and RCP 8.5 (16 % decrease) scenarios. These results indicate that organic carbon outputs and density will decrease faster under high emissions and more slowly under low emissions. This study provides supporting data and a scientific basis for increasing soil organic carbon in cropland and for the management of oasis soil to mitigate the effects of future climate change.

In practical electric vehicle applications, the noise of original discharging/charging voltage (DCV) signals are inevitable, which comes from electromagnetic interference and the measurement noise of the sensors. To solve such problems, the Discrete Wavelet Transform (DWT) based state of charge (SOC) estimation method is proposed in this paper. Through a multi-resolution analysis, the original DCV signals with noise are decomposed into different frequency sub-bands. The desired de-noised DCV signals are then reconstructed by utilizing the inverse discrete wavelet transform, based on the sure rule. With the de-noised DCV signal, the SOC and the parameters are obtained using the adaptive extended Kalman Filter algorithm, and the adaptive forgetting factor recursive least square method. Simulation and experimental results show that the SOC estimation error is less than 1%, which indicates an effective improvement in SOC estimation accuracy.

Pelvic rotation (PR) is commonly seen in patients with idiopathic scoliosis (IS), but factors contributing to this phenomenon and its relationship with the surgical outcome are not well established. This retrospective study included 85 IS patients in 2 groups: thoracic curve dominance group (group A) and lumbar curve dominance group (group B). Pre- and postoperative PR was measured on standing posteroanterior radiographs by the left/right ratio (L/R ratio) of horizontal distance between the anterior superior iliac spine (ASIS) and the inferior ilium (SI) at the sacroiliac joint on the same side in both groups. Other radiographic data, age, sex, and Risser sign of each patient were recorded to analyze their correlations with PR before and after operation. The patients ranged in age from 10 to 35 years with a mean of 17.0 ± 5.2 years. The mean L/R ratio of PR before operation was 0.99 (0.73–1.40) versus 0.98 (0.87–1.26) after operation. The L/R ration was beyond the range of 1 ± 0.1 (indicating the presence of PR) in 17 (20%) patients before operation and in 14 (16.5%) patients after operation. There was no significant difference in PR between the 2 groups of patients either before (P = 0.468) or after (P = 0.944) surgery. The preoperative PR showed a very low correlation with Risser sign (r = 0.220, P = 0.043), apex vertebral rotation (AVR) in the proximal thoracic curve (r = 0.242, P = 0.026), and AVR in the lumbar curve (r = 0.213, P = 0.049), while the postoperative PR showed a very low correlation with Risser sign (r = −0.341, P = 0.001) and postoperative trunk shift (TS) (r = −0.282, P = 0.009). Multiple stepwise regression analysis showed that preoperative PR was affected by proximal thoracic curve AVR and lumbar curve AVR. There was no significant difference between PR before operation and 2 years after operation. Preoperative PR was mainly correlated with Risser sign and the rotation status of the proximal thoracic curve and lumbar curve, while postoperative PR was mainly correlated with Risser sign and postoperative TS.

Jumonji domain-containing protein 1A (JMJD1A) play a key role in the development and progression of several malignancies. The present study investigated the expression and clinical significance of JMJD1A in gastric cancer. JMJD1A was found to be upregulated in gastric cancer tissues and cell lines. Furthermore, the upregulated expression of JMJD1A was significantly correlated with the results of the invasion depth (P=0.006), lymph node metastatic status (P

This study makes an empirical comparison of estimators for censored equations using Monte Carlo simulation. The underlying data generation process is rarely known in practice. From the viewpoint of regression, both ordinary censoring rule and sample selection rule are logical rules of censoring. Furthermore, a mixed censoring rule is also possible to govern underlying data generation process. Therefore, it is valuable to examine whether estimators are robust to variations in the assumptions of censoring rules. Five estimators are examined, estimators for ordinary censoring rules include method of simulated scores, Bayesian estimation, and expectation maximization; estimators for sample selection rules include multivariate Heckman two-step method, and Shonkwiler - Yen two-step method. According to our findings, generally a substantial difference exists in the performance of estimators, and hence the choice of estimator appears to be of importance. Apart from difference in performance, estimates from all procedures are reasonably close to estimated parameters.

This study makes an empirical comparison of estimators for censored equations using Monte Carlo simulation. The underlying data generation process is rarely known in practice. From the viewpoint of regression, both ordinary censoring rule and sample selection rule are logical rules of censoring. Furthermore, a mixed censoring rule is also possible to govern underlying data generation process. Therefore, it is valuable to examine whether estimators are robust to variations in the assumptions of censoring rules. Five estimators are examined, estimators for ordinary censoring rules include method of simulated scores, Bayesian estimation, and expectation maximization; estimators for sample selection rules include multivariate Heckman two-step method, and Shonkwiler - Yen two-step method. According to our findings, generally a substantial difference exists in the performance of estimators, and hence the choice of estimator appears to be of importance. Apart from difference in performance, estimates from all procedures are reasonably close to estimated parameters.

Numerical investigation of the characteristics of an 80° delta wing in combined force-pitch and free-roll is presented. The implicit, upwind, flux-difference splitting, finite volume scheme and the second-order-accurate finite difference scheme are employed to solve the flow governing equations and Euler rigid-body dynamics equations, respectively. The characteristics of the delta wing in combined free-roll and large amplitude force-pitch is obtained numerically and shows a well agreement with experimental data qualitatively. The motion in combined force-pitch and free-roll significantly reduces the lift force and transverse stabilities of the delta wing, which is closely related to the flying safety. Investigations on sensitive factors indicate that the roll-axis moment of inertia and the structural damping have great influence on the frequency and amplitude, respectively. Moreover, the turbulence model is considered as an influencing factor in the investigation., {"references":["R.C. Nelson, A. Pelletier, \"The unsteady aerodynamics of slender wings\nand aircraft undergoing large amplitude maneuvers,\" Progress in\nAerospace Sciences, vol. 39, no.2-3, pp. 185-248. 2003.","L.E. Nguyen, L.P. Yin, and J.R. Chambers, \"Selfinduced wing rock of\nslender delta wing,\" AIAA paper 81-1883, Aug. 1981.","B.N. Pamadi, D.M. Rao, and T. Niranjana, \"Wing rock and roll attractor\nof delta wing at high angles of attack,\" AIAA paper 94-0807, 1994.","Ericson, L., \"Wing rock analysis of slender delta wings, review and\nextension,\" AIAA paper 95-0317, Jan.1995.","X.Z. Huang, and E.S. Hanff, \"Non-linear rolling stability of a 65┬░ delta\nwing model at high incidence,\" AIAA paper 99-4102, 1999.","N.M. Chaderjian, and L.B. Schiff, \"Navier-Stokes prediction of\nlarge-amplitude forced and free-to-roll delta-wing oscillations,\" AIAA\npaper 94-1884, 1994.","O.A. Kandil, and M.A. Menzies, \"Effective control of computationally\nsimulated wing rock in subsonic flow\". AIAA paper 97-0831, 1997.","W. Liu, H.X. Zhang, and H.Y. Zhao, \"Numerical simulation and physical\ncharacteristics analysis for slender wing rock,\" Journal of Aircraft, vol.43,\nno.3, pp. 858-861. 2006.","R.M. Hall, and S.H. Woodson, \"Introduction to the abrupt wing stall\nprogram,\" Journal of Aircraft, vol.41, no.3, pp. 425-435. 2004.\n[10] R.M. Hall, \"Introduction: Abrupt Wing Stall program, Part 2,\" Journal of\nAircraft, vol.42, no.3, pp. 577-577. 2005.\n[11] R.M. Hall, S.H. Woodson, and J.R. Chambers, \"Accomplishments of the\nAbrupt-Wing-Stall program,\" Journal of Aircraft, vol.42, no.3, pp.\n653-660. 2005.\n[12] O.A. Kandil, and H.A. Kandil, \"Pitching oscillation of a 65-degree delta\nwing in transonic vortex-breakdown flow,\" AIAA paper 94-1426-CP,\n1994.\n[13] Y.A. Abdelhamid, and O.A. Kandil, \"Effect of reduced frequency on\nsuper maneuver delta wing,\" AIAA paper 98-0415, 1998.\n[14] C. Jouannet, and P. Krus, \"Lift coefficient predictions for delta wing\nunder pitching motions,\" AIAA paper 2002-2969, 2002.\n[15] O.A. Kandil, and M.A. Menzies, \"Coupled rolling and pitching oscillation\neffects on transonic shock-induced vortex-breakdown flow of a delta\nwing,\" AIAA paper 96-0828, 1996.\n[16] H.J. Kowal, and A.D. Vakili, \"An investigation of unsteady vortex flow\nfor a pitching-rolling 70-deg delta wing,\" AIAA paper98-0416, 1998.\n[17] M.Z. Tang, W. Zhang, and H.L. He, \"Experimental investigation on\nunsteady flow field about a coupled pitching-rolling delta wing,\" Acta\nAerodynamica Sinica, vol. 19, no. 1, pp.47-55. 2001.\n[18] M.J. Khan, and A. Ahmed, \"Response of vortex breakdown induced wing\nrock to pitching & plunging,\" AIAA paper 2004-4732, 2004.\n[19] J.M. Elzebda, D.T. Mook, and A.H. Nayfeh, \"Influence of pitching\nmotion on subsonic wing rock of slender delta wings,\" Journal of Aircraft,\nvol. 26, no. 6, pp. 503-508. 1989.\n[20] J. Er-El, D. Seter, and D. Weihs, \"Nonlinear aerodynamics of a delta wing\nin combined pitch and roll,\" Journal of Aircraft, vol. 26, no. 3, pp.\n245-259. 1989.\n[21] J. Blazek, COMPUTATIONAL FLUID DYNAMICS: PRINCIPLES AND\nAPPLICATIONS (Book style). First edition 2001, ELSEVIER SCIENCE\nLtd. 2001, pp. 16-18.\n[22] Q. Shen, and H.X. Zhang, \"A new upwind NND scheme for Euler\nequations and its application to the supersonic flow,\" in Proceedings of\nAsia Workshop on CFD, Sichuan, China, 1994.\n[23] A. Jameson, \"Time dependent calculations using multigrid with\napplication to unsteady flows past airfoils and wings,\" AIAA paper\n91-1596, 1991."]}