本研究針對 NKN/HfO2 雙層結構電阻式記憶體（RRAM）進行了系統性的研究，重點涵蓋材料結構設計、導電機制分析、類神經特性量測，以及其在人工智慧運算中的應用。隨著物聯網（IoT）、人工智慧（AI）、大數據與邊緣運算的快速發展，傳統記憶體架構逐漸面臨物理限制與高功耗挑戰。RRAM 具備非揮發性、低功耗、結構簡單以及多層次儲存能力，被視為次世代「記憶體與運算整合」（Computing-in-Memory, CIM）架構的關鍵技術。本研究採用無鉛鈣鈦礦材料 NKN，並結合高介電常數材料 HfO2 形成雙層異質結構，期望能兼具多功能性與穩定性，以提升元件的記憶體效能與類神經應用潛力。
在製程部分，本研究利用濺鍍技術沉積 NKN 與 HfO2 薄膜，並形成 Pt/NKN/HfO2/TiN 的 MIM 結構。透過 X 射線光電子能譜（XPS）分析不同製程條件下的化學組成與鍵結狀態，結果顯示隨著 NKN 層厚度增加，Hf 元素擴散至 NKN 晶格的比例上升，伴隨氧空缺濃度增加，進而改善元件的開關電壓與 ON/OFF 比。穿透式電子顯微鏡（TEM）則證實雙層界面的完整性與結晶品質。
在電性量測方面，不同厚度與製程條件的元件進行了 I–V 掃描、耐久性與保持性測試。當 NKN 厚度為 70 nm，HfO2 厚度為 15 nm 時，元件能在多次切換循環下保持穩定，ON/OFF 比大於 200，耐久性可達 10⁵ 次以上且無明顯劣化。此外，隨著 HfO₂ 厚度增加，元件的切換特性進一步改善，推測 NKN/HfO2 界面能形成可控的導電細絲斷裂位置。透過連續脈衝刺激，元件展現了長期增益（LTP）與長期抑制（LTD）行為，證明其具有優異的突觸可塑性與非線性導電調變特徵。同時，當在前突觸與後突觸施加脈衝時，也成功觀察到脈衝時序依賴可塑性（STDP）特性。
在應用模擬部分，本研究將量測到的「脈衝–導電度對應關係」導入硬體感知卷積神經網路（RRAM-aware CNN）與類神經網路（SNN），以評估元件特性對影像辨識效能的影響。在 MNIST 資料集上的實驗顯示，將非線性導電行為與實際硬體限制直接引入訓練過程，可以更真實反映硬體特性，在準確率無顯著降低的情況下達成能耗與硬體對應性的平衡。而在 SNN 架構中，透過 STDP 規則再現了元件在脈衝刺激下的學習行為，並透過視覺化展現脈衝活化分佈與突觸權重演化，驗證了 NKN/HfO2 RRAM 作為人工突觸的可行性。
綜合來看，本研究證實 NKN/HfO2 雙層結構 RRAM 在記憶體與類神經應用領域皆展現優異表現。從材料層級的氧空缺控制，到電性層級的穩定多狀態切換，再到系統層級的硬體感知 AI 模擬，本研究都展現了其作為次世代 CIM 核心元件的潛力。同時，也深化了對鈣鈦礦/高介電異質結構導電機制的理解，並提供了實驗與模擬雙重驗證，為低功耗、高密度、具生物啟發性的計算架構奠定基礎。

This study systematically investigates NKN/HfO2 bilayer resistive random-access memory (RRAM), focusing on material structure design, conduction mechanism analysis, neuromorphic characteristics, and applications in artificial intelligence computing. With the rapid growth of the Internet of Things (IoT), artificial intelligence (AI), big data, and edge computing, traditional memory architectures are facing physical limitations and high power consumption challenges. RRAM, with its non-volatility, low power consumption, simple structure, and multi-level storage capability, has been regarded as a key technology for next-generation Computing-in-Memory (CIM) architectures. In this work, lead-free perovskite NKN combined with a high-k dielectric HfO2 is adopted to form a bilayer heterostructure, aiming to simultaneously achieve multifunctionality and stability, thereby enhancing both memory performance and neuromorphic potential.
In the fabrication process, NKN and HfO2 thin films were deposited via sputtering, forming a Pt/NKN/HfO2/TiN MIM structure. X-ray photoelectron spectroscopy was employed to analyze the chemical composition and bonding states under different processing conditions. Results showed that as the NKN layer thickness increased, the upward diffusion of Hf atoms into the NKN lattice also increased, accompanied by higher oxygen vacancy concentrations. These changes correlated with improvements in switching voltage and ON/OFF ratio. Transmission electron microscopy confirmed the bilayer interface integrity and crystallinity.
For electrical measurements, devices with varying thicknesses and processing conditions were tested by I–V sweeps, endurance cycling, and retention evaluations. When the NKN thickness was 70 nm and the HfO2 buffer layer was 15 nm, the device maintained stable switching behavior over repeated cycles, with an ON/OFF ratio exceeding 200 and endurance up to 10⁵ cycles without significant degradation. Moreover, the device characteristics were found to improve as the HfO2 buffer layer thickness increased, suggesting that the NKNHfO2 interface may provide a controllable rupture point for conductive filaments. By applying consecutive electrical pulses, the device demonstrated long-term potentiation (LTP) and long-term depression (LTD), indicating robust synaptic plasticity consistent with nonlinear conductance modulation. Furthermore, by stimulating pre- and post-synaptic pulses, spike-timing-dependent plasticity (STDP) behavior was successfully observed.
In application-level simulations, the experimentally obtained pulse-to-conductance mapping was integrated into hardware-aware convolutional neural networks (RRAM-aware CNNs) and spiking neural networks (SNNs) to evaluate the influence of device behavior on image recognition performance. On the MNIST dataset, incorporating nonlinear conductance characteristics and device limitations into training enabled more realistic hardware modeling, achieving a balance between accuracy and energy efficiency without significant performance loss. In the SNN framework, STDP rules were applied, successfully reproducing device-based learning dynamics, while spike activity maps and synaptic weight evolution were visualized, validating the feasibility of NKN/HfO2 RRAM as synaptic devices in neuromorphic systems.
In summary, this research demonstrates that NKN/HfO2 bilayer RRAM devices exhibit excellent potential for both memory and neuromorphic applications. Through oxygen vacancy engineering at the material level, stable multi-state switching at the electrical level, and hardware-aware AI simulations at the system level, these devices show great promise as core elements of next-generation CIM architectures. Beyond clarifying the conduction mechanisms in perovskite/high-k heterostructures, this study also provides both experimental and modeling validation toward low-power, high-density, and bio-inspired computing technologies.

[1] W.-S. Khwa, D. Lu, C.-M. Dou, and M.-F. Chang, "Emerging NVM circuit techniques and implementations for energy-efficient systems," in Beyond-CMOS Technologies for Next Generation Computer Design: Springer, 2018, pp. 85-132.
[2] S. Yu, Resistive random access memory (RRAM). Morgan & Claypool Publishers, 2016.
[3] T. Zhang, X. Ou, W. Zhang, J. Yin, Y. Xia, and Z. Liu, "High-k-rare-earth-oxide Eu2O3 films for transparent resistive random access memory (RRAM) devices," Journal of Physics D: Applied Physics, vol. 47, no. 6, p. 065302, 2014.
[4] J. H. Yoon, Y.-W. Song, W. Ham, J.-M. Park, and J.-Y. Kwon, "A review on device requirements of resistive random access memory (RRAM)-based neuromorphic computing," APL Materials, vol. 11, no. 9, 2023.
[5] K. Yeh, P. Shih, K. Hsu, W. Cheng, and S. Chu, "Yttrium Doping Effects on the Resistive Random Access Memory Characteristics of Sputtered HfO x Films and Mechanism Investigations," ACS Applied Electronic Materials, 2025.
[6] X. F. Cheng et al., "Environmentally robust memristor enabled by lead‐free double perovskite for high‐performance information storage," Small, vol. 15, no. 49, p. 1905731, 2019.
[7] D. Yadav, A. K. Dwivedi, S. Verma, and D. K. Avasthi, "Transition Metal Oxide Based Resistive Random-Access Memory: An Overview of Materials and Device Performance Enhancement Techniques," Journal of Science: Advanced Materials and Devices, p. 100813, 2024.
[8] N. Mullani et al., "Improved resistive switching behavior of multiwalled carbon nanotube/TiO2 nanorods composite film by increased oxygen vacancy reservoir," Materials Science in Semiconductor Processing, vol. 108, p. 104907, 2020.
[9] D. G. Jeong et al., "Grain boundary control for high-reliability HfO2-based RRAM," Chaos, Solitons & Fractals, vol. 183, p. 114956, 2024.
[10] B. Santra, G. Das, G. Aquilanti, and A. Kanjilal, "Resistive switching and synaptic characteristics in ZnO@ β-SiC composite-based RRAM for neuromorphic computing," Journal of Applied Physics, vol. 137, no. 4, 2025.
[11] Q. Xue et al., "Transient and Biocompatible Resistive Switching Memory Based on Electrochemically‐Deposited Zinc Oxide," Advanced Electronic Materials, vol. 7, no. 12, p. 2100322, 2021.
[12] S. Roy et al., "Toward a reliable synaptic simulation using Al-doped HfO2 RRAM," ACS applied materials & interfaces, vol. 12, no. 9, pp. 10648-10656, 2020.
[13] C. Zhang, Y. Li, C. Ma, and Q. Zhang, "Recent progress of organic–inorganic hybrid perovskites in RRAM, artificial synapse, and logic operation," Small Science, vol. 2, no. 2, p. 2100086, 2022.
[14] L. Zhang et al., "Advances in the application of perovskite materials," Nano-Micro Letters, vol. 15, no. 1, p. 177, 2023.
[15] L. Ma, M.-G. Ju, J. Dai, and X. C. Zeng, "Tin and germanium based two-dimensional Ruddlesden–Popper hybrid perovskites for potential lead-free photovoltaic and photoelectronic applications," Nanoscale, vol. 10, no. 24, pp. 11314-11319, 2018.
[16] B. Hwang and J.-S. Lee, "Lead-free, air-stable hybrid organic–inorganic perovskite resistive switching memory with ultrafast switching and multilevel data storage," Nanoscale, vol. 10, no. 18, pp. 8578-8584, 2018.
[17] B. Y. Kim et al., "Resistive switching memory integrated with nanogenerator for self‐powered bioimplantable devices," Advanced Functional Materials, vol. 26, no. 29, pp. 5211-5221, 2016.
[18] Y. Yang et al., "Improved resistive switching performance and in-depth mechanism analysis in Mn-doped SrTiO3-based RRAM," Materials Today Communications, vol. 35, p. 105512, 2023.
[19] S. J. Kim et al., "Reliable and Robust Two-Dimensional Perovskite Memristors for Flexible-Resistive Random-Access Memory Array," ACS nano, vol. 18, no. 41, pp. 28131-28141, 2024.
[20] Y. Fang et al., "Improvement of HfO x-based RRAM device variation by inserting ALD TiN buffer layer," IEEE Electron Device Letters, vol. 39, no. 6, pp. 819-822, 2018.
[21] H. He, Y. Tan, C. Lee, and Y. Zhao, "Ti/HfO2-based RRAM with superior thermal stability based on self-limited TiOx," Electronics, vol. 12, no. 11, p. 2426, 2023.
[22] U. B. Isyaku, M. H. B. M. Khir, I. M. Nawi, M. Zakariya, and F. Zahoor, "ZnO based resistive random access memory device: a prospective multifunctional next-generation memory," IEEE Access, vol. 9, pp. 105012-105047, 2021.
[23] H. Kim et al., "MAPbBr3 Halide Perovskite-Based Resistive Random-Access Memories Using Electron Transport Layers for Long Endurance Cycles and Retention Time," ACS Applied Materials & Interfaces, vol. 16, no. 2, pp. 2457-2466, 2024.
[24] W.-C. Chen, S. Qin, Z. Yu, and H.-S. P. Wong, "Reduced HfO₂ Resistive Memory Variability by Inserting a Thin SnO₂ as Oxygen Stopping Layer," IEEE Electron Device Letters, vol. 42, no. 12, pp. 1778-1781, 2021.
[25] H. Nishiyama et al., "Alkali volatilization of (Li, Na, K) NbO3-based piezoceramics and large-field electrical and mechanical properties," Journal of the Ceramic Society of Japan, vol. 129, no. 3, pp. 127-134, 2021.
[26] I. Ahmad, D. Lee, M. Chae, T. Kim, M. Ali, and H.-D. Kim, "Improved resistive switching characteristics observed in amorphous boron nitride-based RRAM device via oxygen doping: A study based on bulk and interface traps analysis," Materials Science in Semiconductor Processing, vol. 184, p. 108805, 2024.
[27] S. Yu, "Overview of resistive switching memory (RRAM) switching mechanism and device modeling," in 2014 IEEE International Symposium on Circuits and Systems (ISCAS), 2014: IEEE, pp. 2017-2020.
[28] D. Zhang, J. Wang, Q. Wu, and Y. Du, "Exploring the direction-dependency of conductive filament formation and oxygen vacancy migration behaviors in HfO 2-based RRAM," Physical Chemistry Chemical Physics, vol. 25, no. 4, pp. 3521-3534, 2023.
[29] Y. Dai, Y. Zhao, J. Wang, J. Xu, and F. Yang, "First principle simulations on the effects of oxygen vacancy in HfO2-based RRAM," AIP Advances, vol. 5, no. 1, 2015.
[30] Y. Nishi and B. Magyari-Kope, Advances in non-volatile memory and storage technology. Woodhead Publishing, 2019.
[31] S. Yu, X. Guan, and H.-S. P. Wong, "Conduction mechanism of TiN/HfOx/Pt resistive switching memory: A trap-assisted-tunneling model," Applied Physics Letters, vol. 99, no. 6, 2011.
[32] E. W. Lim and R. Ismail, "Conduction mechanism of valence change resistive switching memory: A survey," Electronics, vol. 4, no. 3, pp. 586-613, 2015.
[33] O. Abdel-Hamid, L. Deng, and D. Yu, "Exploring convolutional neural network structures and optimization techniques for speech recognition," in Interspeech, 2013, vol. 2013: Citeseer, pp. 1173-5.
[34] F.-C. Chiu, "A review on conduction mechanisms in dielectric films," Advances in Materials Science and Engineering, vol. 2014, no. 1, p. 578168, 2014.
[35] N. F. Mott and E. A. Davis, Electronic processes in non-crystalline materials. OUP Oxford, 2012.
[36] J. Kim et al., "Synaptic characteristics and vector‐matrix multiplication operation in highly uniform and cost‐effective four‐layer vertical RRAM array," Advanced Functional Materials, vol. 34, no. 8, p. 2310193, 2024.
[37] T. Sun et al., "Organic-2D composite material-based RRAM with high reliability for mimicking synaptic behavior," Journal of Materiomics, vol. 10, no. 2, pp. 440-447, 2024.
[38] M. Asif, Y. Singh, A. Thakre, V. Singh, and A. Kumar, "Synaptic plasticity and learning behaviour in multilevel memristive devices," RSC advances, vol. 13, no. 19, pp. 13292-13302, 2023.
[39] D. Ju, M. Noh, S. Lee, J. Lee, S. Kim, and J.-K. Lee, "Investigation of the Versatile Utilization of Three-Dimensional Vertical Resistive Random-Access Memory in Neuromorphic Computing," ACS Applied Materials & Interfaces, vol. 16, no. 43, pp. 59497-59506, 2024.
[40] O. Kwon, J. Shin, D. Chung, and S. Kim, "Energy efficient short-term memory characteristics in Ag/SnOx/TiN RRAM for neuromorphic system," Ceramics International, vol. 48, no. 20, pp. 30482-30489, 2022.
[41] C. Y. Lin et al., "Adaptive synaptic memory via lithium ion modulation in RRAM devices," Small, vol. 16, no. 42, p. 2003964, 2020.
[42] J. Park, M. Kwak, K. Moon, J. Woo, D. Lee, and H. Hwang, "TiO x-based RRAM synapse with 64-levels of conductance and symmetric conductance change by adopting a hybrid pulse scheme for neuromorphic computing," IEEE Electron Device Letters, vol. 37, no. 12, pp. 1559-1562, 2016.
[43] Y. Zhang et al., "Optimized programming scheme Enabling Symmetric conductance Modulation in HfO₂ resistive random-access memory (RRAM) for neuromorphic systems," IEEE Electron Device Letters, vol. 43, no. 8, pp. 1203-1206, 2022.
[44] J. Woo et al., "Improved synaptic behavior under identical pulses using AlO x/HfO 2 bilayer RRAM array for neuromorphic systems," IEEE Electron Device Letters, vol. 37, no. 8, pp. 994-997, 2016.
[45] A. Patnaik, D. Panda, P.-X. Chen, N. Sahoo, and T.-Y. Tseng, "Metal free all oxide SnOx/HfOx bilayer transristor synapse for neuromorphic computing," Journal of Applied Physics, vol. 137, no. 11, 2025.
[46] P.-Y. Chen et al., "Mitigating effects of non-ideal synaptic device characteristics for on-chip learning," in 2015 IEEE/ACM International Conference on Computer-Aided Design (ICCAD), 2015: IEEE, pp. 194-199.
[47] J. Du, "Understanding of object detection based on CNN family and YOLO," in Journal of Physics: Conference Series, 2018, vol. 1004: IOP Publishing, p. 012029.
[48] D. Palaz and R. Collobert, "Analysis of CNN-based speech recognition system using raw speech as input," 2015.
[49] S. E. Kim et al., "Sodium‐doped titania self‐rectifying memristors for crossbar array neuromorphic architectures," Advanced Materials, vol. 34, no. 6, p. 2106913, 2022.
[50] A. Grossi et al., "Fundamental variability limits of filament-based RRAM," in 2016 IEEE International Electron Devices Meeting (IEDM), 2016: IEEE, pp. 4.7. 1-4.7. 4.
[51] S. Yu, B. Gao, Z. Fang, H. Yu, J. Kang, and H.-S. P. Wong, "A neuromorphic visual system using RRAM synaptic devices with sub-pJ energy and tolerance to variability: Experimental characterization and large-scale modeling," in 2012 International Electron Devices Meeting, 2012: IEEE, pp. 10.4. 1-10.4. 4.
[52] L. Chen et al., "Accelerator-friendly neural-network training: Learning variations and defects in RRAM crossbar," in Design, Automation & Test in Europe Conference & Exhibition (DATE), 2017, 2017: IEEE, pp. 19-24.
[53] N. AbuHamra and B. Mohammad, "Memory-Centric Computing for Image Classification Using SNN with RRAM," in 2024 IEEE 6th International Conference on AI Circuits and Systems (AICAS), 2024: IEEE, pp. 105-109.
[54] K. Wong et al., "SNNGX: Securing Spiking Neural Networks with Genetic XOR Encryption on RRAM-based Neuromorphic Accelerator," in Proceedings of the 43rd IEEE/ACM International Conference on Computer-Aided Design, 2024, pp. 1-9.
[55] Y. Zhang, H. Xu, L. Huang, and C. Chen, "A storage-efficient SNN–CNN hybrid network with RRAM-implemented weights for traffic signs recognition," Engineering Applications of Artificial Intelligence, vol. 123, p. 106232, 2023.
[56] J. Chang, V. Sitzmann, X. Dun, W. Heidrich, and G. Wetzstein, "Hybrid optical-electronic convolutional neural networks with optimized diffractive optics for image classification," Scientific reports, vol. 8, no. 1, pp. 1-10, 2018.
[57] A. El Arrassi, A. Gebregiorgis, A. El Haddadi, and S. Hamdioui, "Energy-efficient SNN implementation using RRAM-based computation in-memory (CIM)," in 2022 IFIP/IEEE 30th International Conference on Very Large Scale Integration (VLSI-SoC), 2022: IEEE, pp. 1-6.
[58] Y. Wang et al., "Energy efficient RRAM spiking neural network for real time classification," in Proceedings of the 25th edition on Great Lakes Symposium on VLSI, 2015, pp. 189-194.
[59] S. K. Vohra, S. A. Thomas, M. Sakare, and D. M. Das, "Circuit implementation of on-chip trainable spiking neural network using CMOS based memristive STDP synapses and LIF neurons," Integration, vol. 95, p. 102122, 2024.
[60] W. Gerstner and W. M. Kistler, Spiking neuron models: Single neurons, populations, plasticity. Cambridge university press, 2002.
[61] C. Xiao, J. Chen, and L. Wang, "Optimal mapping of spiking neural network to neuromorphic hardware for edge-AI," Sensors, vol. 22, no. 19, p. 7248, 2022.
[62] S. Bianchi et al., "A compact model for stochastic spike-timing-dependent plasticity (STDP) based on resistive switching memory (RRAM) synapses," IEEE Transactions on Electron Devices, vol. 67, no. 7, pp. 2800-2806, 2020.
[63] Y. Sun, L. Song, L. Hua, W. Cai, W. Chen, and X. Zhao, "Crystal micromorphologies and forming voltage effect on resistance switching behaviors in Ti/Pr (Sr0. 1Ca0. 9) 2Mn2O7/Pt devices," Journal of Alloys and Compounds, vol. 646, pp. 477-482, 2015.
[64] H. Dou et al., "Engineering of grain boundaries in CeO2 enabling tailorable resistive switching properties," Advanced Electronic Materials, vol. 9, no. 5, p. 2201186, 2023.
[65] Y. Taga and R. Takahasi, "Role of kinetic energy of sputtered particles in thin film formation," Surface science, vol. 386, no. 1-3, pp. 231-240, 1997.
[66] C. Cai et al., "Oxygen vacancy formation and uniformity of conductive filaments in Si-doped Ta2O5 RRAM," Applied Surface Science, vol. 560, p. 149960, 2021.
[67] K. Yeh, P. Shih, K. Hsu, W. Cheng, and S. Chu, "Yttrium Doping Effects on the Resistive Random Access Memory Characteristics of Sputtered HfO x Films and Mechanism Investigations," ACS Applied Electronic Materials, vol. 7, no. 5, pp. 1802-1811, 2025.
[68] E. Piros et al., "Role of oxygen defects in conductive-filament formation in Y 2 O 3-based analog RRAM devices as revealed by fluctuation spectroscopy," Physical Review Applied, vol. 14, no. 3, p. 034029, 2020.
[69] L. Zhao, S. Clima, B. Magyari-Köpe, M. Jurczak, and Y. Nishi, "Ab initio modeling of oxygen-vacancy formation in doped-HfOx RRAM: Effects of oxide phases, stoichiometry, and dopant concentrations," Applied Physics Letters, vol. 107, no. 1, 2015.
[70] Y. Ahn and J. Y. Son, "Thickness-dependent resistive switching memory characteristics of NiO nanodisks fabricated by AAO nanotemplate," Current Applied Physics, vol. 54, pp. 44-48, 2023.
[71] J. H. Xi, X. P. Chen, H. X. Li, J. Zhang, and Z. G. Ji, "Effects of Film Thickness on Resistive Switching Characteristics of ZnO Based ReRAM," Advanced Materials Research, vol. 721, pp. 194-198, 2013.
[72] Y. Liu et al., "Effect of film thickness and temperature on the resistive switching characteristics of the Pt/HfO2/Al2O3/TiN structure," Solid-State Electronics, vol. 173, p. 107880, 2020.
[73] T. Kim et al., "Oxide thickness-dependent resistive switching characteristics of Cu/HfO2/Pt ECM devices," Applied Physics Letters, vol. 122, no. 2, 2023.
[74] W. Zhang et al., "Switching-behavior improvement in HfO2/ZnO bilayer memory devices by tailoring of interfacial and microstructural characteristics," Nanotechnology, vol. 33, no. 25, p. 255703, 2022.
[75] S.-Y. Kwon, W.-S. Ko, J.-H. Byun, D.-Y. Lee, H.-D. Lee, and G.-W. Lee, "The Switching Characteristics in Bilayer ZnO/HfO2 Resistive Random-Access Memory, Depending on the Top Electrode," Electronic Materials, vol. 5, no. 2, pp. 71-79, 2024.
[76] H. Ji, Y. Lee, J. Heo, and S. Kim, "Improved resistive and synaptic switching performances in bilayer ZrOx/HfOx devices," Journal of Alloys and Compounds, vol. 962, p. 171096, 2023.
[77] F. M. Puglisi, L. Larcher, G. Bersuker, A. Padovani, and P. Pavan, "An empirical model for RRAM resistance in low-and high-resistance states," IEEE Electron Device Letters, vol. 34, no. 3, pp. 387-389, 2013.
[78] Y. Li et al., "Investigation on the conductive filament growth dynamics in resistive switching memory via a universal Monte Carlo simulator," Scientific reports, vol. 7, no. 1, p. 11204, 2017.
[79] S. Jung et al., "Bottom electrode reactivity and bonding strength effect on resistive switching in HfO2-based RRAM," Materials Science in Semiconductor Processing, vol. 192, p. 109438, 2025.
[80] T. Dehury, S. Kumar, A. S. K. Sinha, M. Gupta, and C. Rath, "Thickness dependent phase transformation and resistive switching performance of HfO2 thin films," Materials Chemistry and Physics, vol. 315, p. 129035, 2024.
[81] Y. Pyo, J.-U. Woo, H.-G. Hwang, S. Nahm, and J. Jeong, "Effect of oxygen vacancy on the conduction modulation linearity and classification accuracy of Pr0. 7Ca0. 3MnO3 memristor," Nanomaterials, vol. 11, no. 10, p. 2684, 2021.
[82] J.-K. Lee, J. Pyo, and S. Kim, "Low-Frequency Noise-Based Mechanism Analysis of Endurance Degradation in Al/αTiOx/Al Resistive Random Access Memory Devices," Materials, vol. 16, no. 6, p. 2317, 2023.
[83] K. M. Kim et al., "Voltage divider effect for the improvement of variability and endurance of TaOx memristor," Scientific reports, vol. 6, no. 1, p. 20085, 2016.
[84] B.-Y. Kim et al., "Nanogenerator-induced synaptic plasticity and metaplasticity of bio-realistic artificial synapses," NPG Asia Materials, vol. 9, no. 5, pp. e381-e381, 2017.
[85] C. F. Stevens and J. F. Wesseling, "Augmentation is a potentiation of the exocytotic process," Neuron, vol. 22, no. 1, pp. 139-146, 1999.
[86] M. Maestro-Izquierdo, M. Gonzalez, and F. Campabadal, "Mimicking the spike-timing dependent plasticity in HfO2-based memristors at multiple time scales," Microelectronic Engineering, vol. 215, p. 111014, 2019.
[87] A. Mao, M. Mohri, and Y. Zhong, "Cross-entropy loss functions: Theoretical analysis and applications," in International conference on Machine learning, 2023: PMLR, pp. 23803-23828.
[88] Y. Tian, D. Su, S. Lauria, and X. Liu, "Recent advances on loss functions in deep learning for computer vision," Neurocomputing, vol. 497, pp. 129-158, 2022.
[89] Q. Wang, Y. Ma, K. Zhao, and Y. Tian, "A comprehensive survey of loss functions in machine learning," Annals of Data Science, vol. 9, no. 2, pp. 187-212, 2022.
[90] Y. LeCun and F. J. Huang, "Loss Functions for Discriminative Training of Energy-Based Models," presented at the Proceedings of the Tenth International Workshop on Artificial Intelligence and Statistics, Proceedings of Machine Learning Research, 2005. [Online]. Available: https://proceedings.mlr.press/r5/lecun05a.html.