Vedant Karia

12 papers C 2Journal 4Unranked 6
YearRankTypeTitle / Venue / Authors
2025 J jnl
Genesis: A Spiking Neuromorphic Accelerator With On-chip Continual Learning.
CoRR
Vedant Karia, Abdullah M. Zyarah, Dhireesha Kudithipudi
2024 — conf
Advancing Neuro-Inspired Lifelong Learning for Edge with Co-Design.
AAAI Spring Symposia
Nicholas Soures, Vedant Karia, Dhireesha Kudithipudi
2024 — conf
PositCL: Compact Continual Learning with Posit Aware Quantization.
ACM Great Lakes Symposium on VLSI
Vedant Karia, Abdullah M. Zyarah, Dhireesha Kudithipudi
2024 J jnl
Probabilistic Metaplasticity for Continual Learning with Memristors.
CoRR
Fatima Tuz Zohora, Vedant Karia, Nicholas Soures, Dhireesha Kudithipudi
2023 J jnl
NeuroBench: Advancing Neuromorphic Computing through Collaborative, Fair and Representative Benchmarking.
CoRR
Jason Yik, Soikat Hasan Ahmed, Zergham Ahmed, Brian Anderson, Andreas G. Andreou, Chiara Bartolozzi, Arindam Basu, Douwe den Blanken, Petrut Bogdan, Sander M. Bohté, Younes Bouhadjar, Sonia M. Buckley, Gert Cauwenberghs, Federico Corradi, Guido de Croon, Andreea Danielescu, Anurag Reddy Daram, Mike Davies, Yigit Demirag, Jason Eshraghian, Jeremy Forest, Steve B. Furber, P. Michael Furlong, Aditya Gilra, Giacomo Indiveri, Siddharth Joshi, Vedant Karia, Lyes Khacef, James C. Knight, Laura Kriener, Rajkumar Kubendran, Dhireesha Kudithipudi, Gregor Lenz, Rajit Manohar, Christian Mayr, Konstantinos P. Michmizos, Dylan R. Muir, Emre Neftci, Thomas Nowotny, Fabrizio Ottati, Ayça Özcelikkale, Noah Pacik-Nelson, Priyadarshini Panda, Pao-Sheng Sun, Melika Payvand, Christian Pehle, Mihai A. Petrovici, Christoph Posch, Alpha Renner, Yulia Sandamirskaya, Clemens JS Schaefer, André van Schaik, Johannes Schemmel, Catherine D. Schuman, Jae-sun Seo, Sumit Bam Shrestha, Manolis Sifalakis, Amos Sironi, Kenneth Michael Stewart, Terrence C. Stewart, Philipp Stratmann, Guangzhi Tang, Jonathan Timcheck, Marian Verhelst, Craig M. Vineyard, Bernhard Vogginger, Amirreza Yousefzadeh, Biyan Zhou, Fatima Tuz Zohora, Charlotte Frenkel, Vijay Janapa Reddi
2022 — conf
ACTION: Automated Hardware-Software Codesign Framework for Low-precision Numerical Format SelecTION in TinyML.
CoNGA
Hamed Fatemi Langroudi, Vedant Karia, Tej Pandit, Becky Mashaido, Dhireesha Kudithipudi
2022 C conf
SCOLAR: A Spiking Digital Accelerator with Dual Fixed Point for Continual Learning.
ISCAS
Vedant Karia, Fatima Tuz Zohora, Nicholas Soures, Dhireesha Kudithipudi
2021 — conf
ALPS: Adaptive Quantization of Deep Neural Networks With GeneraLized PositS.
CVPR Workshops
Hamed Fatemi Langroudi, Vedant Karia, Zachariah Carmichael, Abdullah M. Zyarah, Tej Pandit, John L. Gustafson, Dhireesha Kudithipudi
2021 C conf
MetaplasticNet: Architecture with Probabilistic Metaplastic Synapses for Continual Learning.
ISCAS
Fatima Tuz Zohora, Vedant Karia, Anurag Reddy Daram, Abdullah M. Zyarah, Dhireesha Kudithipudi
2021 J jnl
TENT: Efficient Quantization of Neural Networks on the tiny Edge with Tapered FixEd PoiNT.
CoRR
Hamed Fatemi Langroudi, Vedant Karia, Tej Pandit, Dhireesha Kudithipudi
2020 — conf
Adaptive Posit: Parameter aware numerical format for deep learning inference on the edge.
CVPR Workshops
Hamed Fatemi Langroudi, Vedant Karia, John L. Gustafson, Dhireesha Kudithipudi
2018 — conf
Scalable IP Core for Feed Forward Random Networks.
ELM
Anurag Reddy Daram, Karan Paluru, Vedant Karia, Dhireesha Kudithipudi
test1.txt
← Index test1.txt plaintext
Truncated preview: showing the first 16384 bytes out of 43286155 total.
{'decompiled': [{'decompiled_function_hash': '1f0dd5f0fe1f0f28d39c8f98bf5eb5b06a98405389873053efd1df2f3ca12179', 'decompiled_function': 'bool z\n\nif (arg3 == *arg2)\n    *arg2 = arg1\n    z = true\nelse\n    *arg2\n    z = false\n\nint64_t result\nresult.b = z\nreturn result', 'decompiled_function_name': 'sync/atomic.CompareAndSwapUintptr', 'decompiled_function_prototype': 'int64_t sync/atomic.CompareAndSwapUintptr(int64_t arg1, int64_t* arg2 @ rax, int64_t arg3 @ rbx)', 'decompiled_function_address': 4198528, 'function_type': 'USER', 'functions_caller': [], 'functions_call': [], 'flattened_score': 0.0, 'mba_score': 0.0, 'disassembled_function_hash': '09e03a03d3a93deea76172fc49c83d407cff050807a0e40d5b25b8848182f060'}, {'decompiled_function_hash': '02b054c1841b6532a667e3e61f8c3363dffbc0893dc9cee31e16e6438a6b32ee', 'decompiled_function': 'int64_t rcx = data_6531e8\nint64_t* internal/abi.kindNames_1 = internal/abi.kindNames\n\nif (arg1 s>= rcx)\n    if (rcx u> 0)\n        internal/abi.kindNames_1[1]\n        return *internal/abi.kindNames_1\n    \n    runtime.panicIndex(0)\nelse if (arg1 u< rcx)\n    internal/abi.kindNames_1[arg1 * 2 + 1]\n    return internal/abi.kindNames_1[arg1 * 2]\n\nruntime.panicIndexU()\nbreakpoint', 'decompiled_function_name': 'internal/abi.Kind.String', 'decompiled_function_prototype': 'int64_t internal/abi.Kind.String(int64_t arg1 @ rax)', 'decompiled_function_address': 4198720, 'function_type': 'USER', 'functions_caller': [], 'functions_call': ['runtime.panicIndex(0)', 'runtime.panicIndexU()'], 'flattened_score': 0.0, 'mba_score': 0.0, 'disassembled_function_hash': 'ba0c8a4a2a408f300d085876a7263294d4b54f4214bc693afe46b9e8378cbf31'}, {'decompiled_function_hash': '03a962adee69ce6e97437b8250ca1e912d9f55f0c17a29845e7e55e9544b5fa9', 'decompiled_function': 'if ((zx.q(*(arg1 + 0x17)) & 0x1f) != 0x11)\n    return 0\n\nreturn *(arg1 + 0x40)', 'decompiled_function_name': 'internal/abi.(*Type).Len', 'decompiled_function_prototype': 'int64_t internal/abi.(*Type).Len(void* arg1 @ rax)', 'decompiled_function_address': 4198816, 'function_type': 'USER', 'functions_caller': [], 'functions_call': [], 'flattened_score': 0.0, 'mba_score': 0.3333333333333333, 'disassembled_function_hash': 'a2843f9a81f7822d6b6b55e43df329ee31eed92dae8dc41ca991d56d7632c1b9'}, {'decompiled_function_hash': 'd782b427b58ad2cf84c6a7c609d1d872f14db0788919434f18e9c183125af6d3', 'decompiled_function': 'void* rsi_4\n\nif ((*(arg1 + 0x14) & 1) == 0)\n    rsi_4 = nullptr\nelse\n    int64_t rsi_3 = (zx.q(*(arg1 + 0x17)) & 0x1f) - 0x11\n    \n    if (rsi_3 u<= 8)\n        jump(internal/abi.(*Type).ExportedMethods.jump7[rsi_3])\n    \n    rsi_4 = arg1 + 0x30\n\nif (rsi_4 == 0)\n    return 0\n\nuint64_t rdx = zx.q(*(rsi_4 + 6))\n\nif (rdx.w == 0)\n    return nullptr\n\nif (rdx u<= 0x10000)\n    return rsi_4 + zx.q(*(rsi_4 + 8))\n\nruntime.panicSlice3Alen()\nbreakpoint', 'decompiled_function_name': 'internal/abi.(*Type).ExportedMethods', 'decompiled_function_prototype': 'void* internal/abi.(*Type).ExportedMethods(void* arg1 @ rax)', 'decompiled_function_address': 4198880, 'function_type': 'USER', 'functions_caller': [], 'functions_call': ['jump(internal/abi.(*Type).ExportedMethods.jump7[rsi_3])', 'runtime.panicSlice3Alen()'], 'flattened_score': 0.0, 'mba_score': 0.125, 'disassembled_function_hash': 'e2558c9cf3ebac3a28dff29dcc0b8918965ca5091e4d5e0361af301ebc9d7fba'}, {'decompiled_function_hash': '407d3d24d10ffe65a17c09aa88e6d7283d00fcdd9df10741af968c91841cc9b2', 'decompiled_function': 'while (&__return_addr u<= *(arg3 + 0x10))\n    runtime.morestack_noctxt.abi0()\n\nuint64_t rcx_1 = zx.q(*(arg1 + 0x17)) & 0x1f\n\nif (rcx_1 == 0x14)\n    return *(arg1 + 0x40)\n\ninternal/abi.(*Type).ExportedMethods(arg1, rcx_1)\nreturn arg2', 'decompiled_function_name': 'internal/abi.(*Type).NumMethod', 'decompiled_function_prototype': 'int64_t internal/abi.(*Type).NumMethod(void* arg1 @ rax, int64_t arg2 @ rbx, void* arg3 @ r14)', 'decompiled_function_address': 4199072, 'function_type': 'USER', 'functions_caller': [], 'functions_call': ['internal/abi.(*Type).ExportedMethods(arg1, rcx_1)', 'runtime.morestack_noctxt.abi0()'], 'flattened_score': 1.0, 'mba_score': 0.14285714285714285, 'disassembled_function_hash': '270037f0c8addf478f25bc075eb0dc05fc54dff87fba6dff63ab8d0555e4929a'}, {'decompiled_function_hash': '9f2a44ad4622a0e2d0e1b439b1d46754301fd5f7060b91e3766482b1383494da', 'decompiled_function': 'if (arg1 == 0)\n    return 0\n\nint64_t rcx = 0\nint64_t rdx = 0\n\nwhile (true)\n    int64_t rbx_2 = rcx * 7\n    *(arg1 + rcx + 1)\n    \n    if (rbx_2 s< 0)\n        runtime.panicshift()\n        breakpoint\n    \n    uint32_t rdi_3 = zx.d(*(arg1 + rcx + 1))\n    rdx += (zx.q(rdi_3) & 0x7f) << rbx_2.b & sbb.q(rbx_2, rbx_2, rbx_2 u< 0x40)\n    \n    if ((rdi_3.b & 0x80) == 0)\n        break\n    \n    rcx += 1\n\nint64_t rcx_1\n\nif (rdx != 1)\n    rcx_1 = 0\nelse\n    rcx_1.b = *(arg1 + 2) == 0x5f\n\nreturn zx.q(rcx_1.d)', 'decompiled_function_name': 'internal/abi.Name.IsBlank', 'decompiled_function_prototype': 'uint64_t internal/abi.Name.IsBlank(void* arg1 @ rax)', 'decompiled_function_address': 4199168, 'function_type': 'USER', 'functions_caller': [], 'functions_call': ['runtime.panicshift()'], 'flattened_score': 0.7692307692307693, 'mba_score': 0.05, 'disassembled_function_hash': '7a693574e5a54de381301402832a96b61c8f4d2a2671d8150e93fb29f8692e0f'}, {'decompiled_function_hash': 'aaadbfaff6f26ed203060269065749f25f44d295aadde2386ab73f8c4a5dd767', 'decompiled_function': 'while (true)\n    int64_t* rsp\n    \n    if (rsp u> *(arg3 + 0x10))\n        *(rsp - 8) = arg2\n        rsp -= 8\n        arg2 = rsp\n        \n        if (arg1 == 0)\n            *rsp\n            return 0\n        \n        void* rcx_1 = nullptr\n        int64_t rdx_1 = 0\n        \n        while (true)\n            void* rsi_2 = rcx_1 * 7\n            *(rcx_1 + arg1 + 1)\n            \n            if (rsi_2 s< 0)\n                break\n            \n            uint32_t r8_3 = zx.d(*(rcx_1 + arg1 + 1))\n            rdx_1 += (zx.q(r8_3) & 0x7f) << rsi_2.b & sbb.q(rsi_2, rsi_2, rsi_2 u< 0x40)\n            \n            if ((r8_3.b & 0x80) == 0)\n                if (rdx_1 s>= 0)\n                    int64_t rcx_4 = neg.q(rcx_1 + arg1 + 2)\n                    \n                    if (rcx_4 u>= rdx_1)\n                        *rsp\n                        return rcx_1 + arg1 + 2\n                    \n                    if (rcx_1 + arg1 != -2)\n                        runtime.panicunsafestringlen(rcx_4)\n                    \n                    runtime.panicunsafestringnilptr()\n                \n                runtime.panicunsafestringlen()\n                break\n            \n            rcx_1 += 1\n        \n        arg1 = runtime.panicshift()\n    \n    rsp[1] = arg1\n    runtime.morestack_noctxt.abi0()\n    arg1 = rsp[1]', 'decompiled_function_name': 'internal/abi.Name.Name', 'decompiled_function_prototype': 'void* internal/abi.Name.Name(int64_t arg1 @ rax, int64_t* arg2 @ rbp, void* arg3 @ r14)', 'decompiled_function_address': 4199296, 'function_type': 'USER', 'functions_caller': [], 'functions_call': ['runtime.panicunsafestringlen(rcx_4)', 'runtime.panicunsafestringnilptr()', 'runtime.panicunsafestringlen()', 'arg1 = runtime.panicshift()', 'runtime.morestack_noctxt.abi0()'], 'flattened_score': 1.0, 'mba_score': 0.029411764705882353, 'disassembled_function_hash': '21ad7fe65b5907f2e2bd82711c3d5cda50e2cda11dd6cb13e472a1879a0f5ee0'}, {'decompiled_function_hash': '6bc616921202fbee6edcbe36548ca6c0123ae51ecb86c0eea217f3889c581d25', 'decompiled_function': 'while (true)\n    int64_t* rsp\n    \n    if (rsp u> *(arg3 + 0x10))\n        *(rsp - 8) = arg2\n        rsp -= 8\n        arg2 = rsp\n        \n        if ((*arg1 & 2) == 0)\n            *rsp\n            return 0\n        \n        void* rcx_1 = nullptr\n        void* rdx_2 = nullptr\n        \n        while (true)\n            void* rsi_2 = rcx_1 * 7\n            *(rcx_1 + arg1 + 1)\n            \n            if (rsi_2 s< 0)\n                break\n            \n            uint32_t r8_3 = zx.d(*(rcx_1 + arg1 + 1))\n            rdx_2 += (zx.q(r8_3) & 0x7f) << rsi_2.b & sbb.q(rsi_2, rsi_2, rsi_2 u< 0x40)\n            \n            if ((r8_3.b & 0x80) == 0)\n                int64_t rcx_3 = 0\n                int64_t rbx_3 = 0\n                \n                while (true)\n                    void* rsi_4 = rdx_2 + rcx_1 + rcx_3\n                    int64_t rdi_3 = rcx_3 * 7\n                    *(arg1 + rsi_4 + 2)\n                    \n                    if (rdi_3 s< 0)\n                        break\n                    \n                    uint32_t r8_9 = zx.d(*(arg1 + rsi_4 + 2))\n                    rbx_3 += (zx.q(r8_9) & 0x7f) << rdi_3.b & sbb.q(rdi_3, rdi_3, rdi_3 u< 0x40)\n                    \n                    if ((r8_9.b & 0x80) == 0)\n                        if (rbx_3 s>= 0)\n                            int64_t rcx_6 = neg.q(arg1 + rsi_4 + 3)\n                            \n                            if (rcx_6 u>= rbx_3)\n                                *rsp\n                                return arg1 + rsi_4 + 3\n                            \n                            if (arg1 + rsi_4 != -3)\n                                runtime.panicunsafestringlen(rcx_6)\n                            \n                            runtime.panicunsafestringnilptr()\n                        \n                        runtime.panicunsafestringlen()\n                        break\n                    \n                    rcx_3 += 1\n                \n                runtime.panicshift()\n                break\n            \n            rcx_1 += 1\n        \n        arg1 = runtime.panicshift()\n    \n    rsp[1] = arg1\n    runtime.morestack_noctxt.abi0()\n    arg1 = rsp[1]', 'decompiled_function_name': 'internal/abi.Name.Tag', 'decompiled_function_prototype': 'void* internal/abi.Name.Tag(char* arg1 @ rax, int64_t* arg2 @ rbp, void* arg3 @ r14)', 'decompiled_function_address': 4199488, 'function_type': 'USER', 'functions_caller': [], 'functions_call': ['runtime.panicunsafestringlen(rcx_6)', 'runtime.panicunsafestringnilptr()', 'runtime.panicunsafestringlen()', 'runtime.panicshift()', 'arg1 = runtime.panicshift()', 'runtime.morestack_noctxt.abi0()'], 'flattened_score': 1.0, 'mba_score': 0.041666666666666664, 'disassembled_function_hash': '0621c0c17af04d429b48b7855fd82627f2e835ec864eaa49d0de3183fd3d9fbf'}, {'decompiled_function_hash': '917d57a1b59aad9ee7b1197909e67f4ee8d39a0d26f2390ceacbbc54bde37a43', 'decompiled_function': 'while (true)\n    void* rsp\n    \n    if (rsp - 0x30 u> *(arg9 + 0x10))\n        *(rsp - 8) = arg6\n        arg6 = rsp - 8\n        rsp -= 0xb0\n        *(rsp + 0xb8) = arg4\n        *(rsp + 0xc8) = arg1\n        \n        if (arg5 s< 0x20000000)\n            if (arg8 s< 0x20000000)\n                __builtin_memset(rsp + 0x84, 0, 0x14)\n                int64_t rdx = arg5\n                int64_t r9_1 = 0\n                \n                while (true)\n                    char r10_2 = arg5.b\n                    arg5 s>>= 7\n                    char r10_3 = r10_2 & 0x7f\n                    \n                    if (arg5 == 0)\n                        if (r9_1 u< 0xa)\n                            *(rsp + r9_1 + 0x8e) = r10_3\n                            int64_t rbx = arg8\n                            int64_t r10_4 = 0\n                            \n                            while (true)\n                                char r11_2 = arg8.b\n                                arg8 s>>= 7\n                                char r11_3 = r11_2 & 0x7f\n                                \n                                if (arg8 == 0)\n                                    if (r10_4 u< 0xa)\n                                        *(rsp + r10_4 + 0x84) = r11_3\n                                        int64_t r11_4 = rdx + r9_1\n                                        int64_t r12_3\n                                        \n                                        if (rbx == 0)\n                                            r12_3 = r9_1 + rdx + 2\n                                        else\n                                            r12_3 = rbx + r10_4 + r11_4 + 3\n                                            arg7 |= 2\n                                        \n                                        *(rsp + 0x78) = rdx\n                                        *(rsp + 0x48) = r9_1 + rdx + 2\n                                        *(rsp + 0x70) = rbx\n                                        *(rsp + 0xb8) = arg4\n                                        *(rsp + 0xc8) = arg1\n                                        *(rsp + 0x68) = r11_4\n                                        *(rsp + 0x50) = r10_4\n                                        *(rsp + 0x58) = r9_1\n                                        \n                                        if (arg3 != 0)\n                                            arg7 |= 8\n                                        \n                                        *(rsp + 0x3f) = arg7.b\n                                        *(rsp + 0x60) = r12_3\n                                        char* result = runtime.makeslice(r12_3)\n                                        void* rcx_1 = *(rsp + 0x60)\n                                        \n                                        if (rcx_1 u> 0)\n                                            *(rsp + 0xa0) = result\n                                            *result = *(rsp + 0x3f)\n                                            void* rsi = rcx_1 - 1\n                                            uint64_t rdx_6 = (zx.q((neg.q(rcx_1 - 1) s>> 0x3f).d) & 1) + result\n                                            void* rdi_2 = *(rsp + 0x58)\n                                            \n                                            if (rdi_2 + 1 s< rsi)\n                                                rsi = rdi_2 + 1\n                                            \n                                            if (rsp + 0x8e != rdx_6)\n                                                runtime.memmove(rsi, rdx_6, rsp + 0x8e)\n                                                result = *(rsp + 0xa0)\n                                                rcx_1 = *(rsp + 0x60)\n                                                rdi_2 = *(rsp + 0x58)\n                                            \n                                            if (rcx_1 u>= rdi_2 + 2)\n                                                void* rdi_3 = rcx_1 - rdi_2 - 2\n                                                void* rdx_9 = ((rdi_2 + 2) & neg.q(rcx_1 - rdi_2 - 2) s>> 0x3f) + result\n                                                void* rsi_6 = *(rsp + 0x78)\n                                                \n                                                if (rsi_6 s< rdi_3)\n                                                    rdi_3 = rsi_6\n                                                \n                                                int128_t* rbx_3 = *(rsp + 0xb8)\n                                                void* rdx_10\n                                                bool cond:0_1\n                                                \n                                                if (rbx_3 != rdx_9)\n                                                    runtime.memmove(rdi_3, rdx_9, rbx_3)\n                                                    rdx_10 = *(rsp + 0x70)\n                                                    cond:0_1 = rdx_10 == 0\n                                                    result = *(rsp + 0xa0)\n                                                    rcx_1 = *(rsp + 0x60)\n                                                else\n                                                    rdx_10 = *(rsp + 0x70)\n                                                    cond:0_1 = rdx_10 == 0\n                                                \n                                                if (cond:0_1)\n                                                    *(rsp + 0xa8)\n                                                    return result\n                                                \n                                                uint64_t rsi_7 = *(rsp + 0x48)\n