Steven J. Byrnes
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Steven J. Byrnes has a total of 71 co-authors. Here are the top co-authors by number of shared publications:
- Federico Capasso
John A. Paulson School of Engineering and Applied Sciences, Harvard University, Cambridge, USA
5 shared publications - Chan-Ho Yang
Department of Physics, Korea Advanced Institute of Science and Technology, Daejeon 34141, Republic of Korea
2 shared publications - Joel Ager
Lawrence Berkeley National Laboratory
2 shared publications L. W. Martin
Materials Sciences Division, Lawrence Berkeley National Laboratory, Berkeley, 94720, California, USA
2 shared publicationsR. Ramesh
Department of Materials Science and Engineering, University of California, Berkeley, Berkeley, 94720, California, USA
2 shared publicationsRomain Blanchard
School of Engineering and Applied Sciences, Harvard University, Cambridge, MA 02138
2 shared publications
Latest Publications
- Open Access
- 13 July 2026
- Published by Center for Open Science in OSF Preprints
Intro to Brain-Like-AGI Safety
Suppose we someday build an Artificial General Intelligence (AGI) algorithm using similar principles of learning and cognition as the human brain. How would we use such an algorithm safely? I argue that this is an open technical problem, and my goal is to bring readers with no prior knowledge all the way up to the front-line of unsolved problems. Chapter 1 has background and motivation; Chapters 2–7 are on neuroscience, arguing for a picture of the brain that combines large-scale learning algorithms (e.g. in the cortex) and specific evolved reflexes (e.g. in the hypothalamus and brainstem); and Chapters 8–15 apply those neuroscience ideas to AGI safety. A major theme is the idea that the brain has something like a reinforcement learning reward function, which says that pain is bad, eating-when-hungry is good, etc. I argue that this reward function is centered around the hypothalamus and brainstem, and that all human desires—even “higher” desires for things like compassion and justice—come directly or indirectly from that innate reward function. If future programmers build brain-like AGI, they will likewise have a reward function slot in their source code, in which they can put whatever they want. If they put the wrong thing, the resulting AGI will wind up callously indifferent to human welfare. How might they avoid that? That’s an open technical problem, but I will review some ideas and research directions.
- 13 July 2026
- Published by Center for Open Science in OSF Preprints
Open AccessIntro to Brain-Like-AGI Safety
Suppose we someday build an Artificial General Intelligence (AGI) algorithm using similar principles of learning and cognition as the human brain. How would we use such an algorithm safely? I argue that this is an open technical problem, and my goal is to bring readers with no prior knowledge all the way up to the front-line of unsolved problems. Chapter 1 has background and motivation; Chapters 2–7 are on neuroscience, arguing for a picture of the brain that combines large-scale learning algorithms (e.g. in the cortex) and specific evolved reflexes (e.g. in the hypothalamus and brainstem); and Chapters 8–15 apply those neuroscience ideas to AGI safety. A major theme is the idea that the brain has something like a reinforcement learning reward function, which says that pain is bad, eating-when-hungry is good, etc. I argue that this reward function is centered around the hypothalamus and brainstem, and that all human desires—even “higher” desires for things like compassion and justice—come directly or indirectly from that innate reward function. If future programmers build brain-like AGI, they will likewise have a reward function slot in their source code, in which they can put whatever they want. If they put the wrong thing, the resulting AGI will wind up callously indifferent to human welfare. How might they avoid that? That’s an open technical problem, but I will review some ideas and research directions.
- 30 June 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE) in Journal of Microelectromechanical Systems
- Vol. 29 (5) , 1008-1013
Polysilicon Grating Switches for LiDAR
Novel polysilicon grating switches are demonstrated, consisting of an optical grating suspended on a moving platform above a waveguide. Electrostatic attraction pulls the grating into close proximity to the waveguide, ejecting light out of the waveguide plane. The switches move only 0.9 to 1.4 μm, and have MHz mechanical resonant frequencies, making them robust to mechanical vibration and shock. These switches enable a new mode of LiDAR imaging; by creating a focal plane array of switches, and coupling it with a standard camera lens, the LiDAR beam can be scanned across the scene with high speed and arbitrary pattern. Both time of flight and coherent ranging techniques are compatible with this approach. We have demonstrated a 10x 10 pixel array of these switches in a LiDAR system, collecting 3D raging data at up to 15 meters away and 5 frames per second, over a 1 degree field of view with 0.1 degree resolution. Additionally, we have shown the switches to be robust, efficient, and reliable, actuating at 15 Volts, with speeds faster than 4 μs, for over 9 billion cycles without degradation or failure. [2020-0132]
- 30 June 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE) in Journal of Microelectromechanical Systems
- Vol. 29 (5) , 1008-1013
Polysilicon Grating Switches for LiDAR
Novel polysilicon grating switches are demonstrated, consisting of an optical grating suspended on a moving platform above a waveguide. Electrostatic attraction pulls the grating into close proximity to the waveguide, ejecting light out of the waveguide plane. The switches move only 0.9 to 1.4 μm, and have MHz mechanical resonant frequencies, making them robust to mechanical vibration and shock. These switches enable a new mode of LiDAR imaging; by creating a focal plane array of switches, and coupling it with a standard camera lens, the LiDAR beam can be scanned across the scene with high speed and arbitrary pattern. Both time of flight and coherent ranging techniques are compatible with this approach. We have demonstrated a 10x 10 pixel array of these switches in a LiDAR system, collecting 3D raging data at up to 15 meters away and 5 frames per second, over a 1 degree field of view with 0.1 degree resolution. Additionally, we have shown the switches to be robust, efficient, and reliable, actuating at 15 Volts, with speeds faster than 4 μs, for over 9 billion cycles without degradation or failure. [2020-0132]
- 10 January 2020
- Published by Optica Publishing Group in Optics Express
- Vol. 28 (2) , 1585-1594
Surface-emitting electroholographic SAW modulator
We report the design and operation of a surface-emitting surface acoustic wave (SAW) acousto-optical modulator which behaves as a cm-scale linear hologram in response to an applied electronic waveform. The modulator is formed by an optical waveguide, transducer, and out-coupling surface grating on a 1 mm-thick lithium niobate substrate. We demonstrate the ability to load and illuminate a 9-region linear hologram into the modulator's 8 mm-long interaction region using applied waveforms of 280–320 MHz. To the best of the authors’ knowledge, this is the first demonstration of a monolithically-integrated, surface-emitting SAW modulator fabricated using lithographic techniques. Applications include practical implementations of a holographic display.
- 10 January 2020
- Published by Optica Publishing Group in Optics Express
- Vol. 28 (2) , 1585-1594
Surface-emitting electroholographic SAW modulator
We report the design and operation of a surface-emitting surface acoustic wave (SAW) acousto-optical modulator which behaves as a cm-scale linear hologram in response to an applied electronic waveform. The modulator is formed by an optical waveguide, transducer, and out-coupling surface grating on a 1 mm-thick lithium niobate substrate. We demonstrate the ability to load and illuminate a 9-region linear hologram into the modulator's 8 mm-long interaction region using applied waveforms of 280–320 MHz. To the best of the authors’ knowledge, this is the first demonstration of a monolithically-integrated, surface-emitting SAW modulator fabricated using lithographic techniques. Applications include practical implementations of a holographic display.
- Open Access
- 12 April 2018
- Published by Optica Publishing Group in Applied Optics
- Vol. 57 (12) , 2973-2980
Substrate aberration and correction for meta-lens imaging: an analytical approach
Meta-lenses based on flat optics enabled a fundamental shift in lens production—providing an easier manufacturing process with an increase in lens profile precision and a reduction in size and weight. Here we present an analytical approach to correct spherical aberrations caused by light propagation through the substrate by adding a substrate-corrected phase profile, which differs from the original hyperbolic one. A meta-lens encoding the new phase profile would yield diffraction-limited focusing and an increase of up to 0.3 of its numerical aperture without changing the radius or focal length. In tightly focused laser spot applications such as direct laser lithography and laser printing, a substrate-corrected meta-lens can reduce the spatial footprint of the meta-lens.
- 12 April 2018
- Published by Optica Publishing Group in Applied Optics
- Vol. 57 (12) , 2973-2980
Open AccessSubstrate aberration and correction for meta-lens imaging: an analytical approach
Meta-lenses based on flat optics enabled a fundamental shift in lens production—providing an easier manufacturing process with an increase in lens profile precision and a reduction in size and weight. Here we present an analytical approach to correct spherical aberrations caused by light propagation through the substrate by adding a substrate-corrected phase profile, which differs from the original hyperbolic one. A meta-lens encoding the new phase profile would yield diffraction-limited focusing and an increase of up to 0.3 of its numerical aperture without changing the radius or focal length. In tightly focused laser spot applications such as direct laser lithography and laser printing, a substrate-corrected meta-lens can reduce the spatial footprint of the meta-lens.
- Open Access
- 2 August 2016
- Published by Optica Publishing Group in Optics Express
- Vol. 24 (16) , 18399-18407
High-quality-factor planar optical cavities with laterally stopped, slowed, or reversed light
In a planar optical cavity, the resonance frequencies increase as a function of in-plane wavevector according to a standard textbook formula. This has well-known consequences in many different areas of optics, from the shifts of etalon peaks at non-normal angles, to the properties of transverse modes in laser diodes, to the effective mass of microcavity photons, and so on. However, this standard formula is valid only when the reflection phase of each cavity mirror is approximately independent of angle. There is a certain type of mirror—a subwavelength dielectric grating near a guided mode resonance—with not only a strongly angle-dependent reflection phase, but also very high reflectance and low losses. Simulations show that by using such mirrors, high-quality-factor planar cavities can be designed that break all these textbook rules, leading to resonant modes that are slow, stopped or even backward-propagating in the in-plane direction. In particular, we demonstrate experimentally high-Q planar cavities whose resonance frequency is independent of in-plane wavevector—i.e., the resonant modes have zero in-plane group velocity, for one polarization but both in-plane directions. We discuss potential applications in various fields including lasers, quantum optics, and exciton-polariton condensation.
- 2 August 2016
- Published by Optica Publishing Group in Optics Express
- Vol. 24 (16) , 18399-18407
Open AccessHigh-quality-factor planar optical cavities with laterally stopped, slowed, or reversed light
In a planar optical cavity, the resonance frequencies increase as a function of in-plane wavevector according to a standard textbook formula. This has well-known consequences in many different areas of optics, from the shifts of etalon peaks at non-normal angles, to the properties of transverse modes in laser diodes, to the effective mass of microcavity photons, and so on. However, this standard formula is valid only when the reflection phase of each cavity mirror is approximately independent of angle. There is a certain type of mirror—a subwavelength dielectric grating near a guided mode resonance—with not only a strongly angle-dependent reflection phase, but also very high reflectance and low losses. Simulations show that by using such mirrors, high-quality-factor planar cavities can be designed that break all these textbook rules, leading to resonant modes that are slow, stopped or even backward-propagating in the in-plane direction. In particular, we demonstrate experimentally high-Q planar cavities whose resonance frequency is independent of in-plane wavevector—i.e., the resonant modes have zero in-plane group velocity, for one polarization but both in-plane directions. We discuss potential applications in various fields including lasers, quantum optics, and exciton-polariton condensation.
- Open Access
- 1 March 2016
- Published by Optica Publishing Group in Optics Express
- Vol. 24 (5) , 5110-5124
Designing large, high-efficiency, high-numerical-aperture, transmissive meta-lenses for visible light
A metasurface lens (meta-lens) bends light using nanostructures on a flat surface. Macroscopic meta-lenses (mm- to cm-scale diameter) have been quite difficult to simulate and optimize, due to the large area, the lack of periodicity, and the billions of adjustable parameters. We describe a method for designing a large-area meta-lens that allows not only prediction of the efficiency and far-field, but also optimization of the shape and position of each individual nanostructure, with a computational cost that is almost independent of the lens size. As examples, we design three large NA = 0.94 meta-lenses: One with 79% predicted efficiency for yellow light, one with dichroic properties, and one broadband lens. All have a minimum feature size of 100nm.
- 1 March 2016
- Published by Optica Publishing Group in Optics Express
- Vol. 24 (5) , 5110-5124
Open AccessDesigning large, high-efficiency, high-numerical-aperture, transmissive meta-lenses for visible light
A metasurface lens (meta-lens) bends light using nanostructures on a flat surface. Macroscopic meta-lenses (mm- to cm-scale diameter) have been quite difficult to simulate and optimize, due to the large area, the lack of periodicity, and the billions of adjustable parameters. We describe a method for designing a large-area meta-lens that allows not only prediction of the efficiency and far-field, but also optimization of the shape and position of each individual nanostructure, with a computational cost that is almost independent of the lens size. As examples, we design three large NA = 0.94 meta-lenses: One with 79% predicted efficiency for yellow light, one with dichroic properties, and one broadband lens. All have a minimum feature size of 100nm.
- 3 March 2014
- Published by National Academy of Sciences in Proceedings of the National Academy of Sciences
- Vol. 111 (11) , 3927-3932
Harvesting renewable energy from Earth’s mid-infrared emissions
Significance: Renewable energy can be generated whenever heat flows from a hotter to a colder body. One such flow is from the warm surface of Earth to cold outer space, via infrared thermal radiation. An emissive energy harvester (EEH) is a device that can generate energy from emitting thermal radiation into the clear sky. We calculate how much power is thermodynamically available, using a location in Oklahoma as a case study. We discuss two possible ways to make such a device: A thermal EEH (analogous to solar thermal power generation) and an optoelectronic EEH (analogous to photovoltaic power generation). For the latter, we propose using a rectifying antenna, and we discuss its operating principles, efficiency limits, system design considerations, and possible technological implementations.
- 3 March 2014
- Published by National Academy of Sciences in Proceedings of the National Academy of Sciences
- Vol. 111 (11) , 3927-3932
Harvesting renewable energy from Earth’s mid-infrared emissions
Significance: Renewable energy can be generated whenever heat flows from a hotter to a colder body. One such flow is from the warm surface of Earth to cold outer space, via infrared thermal radiation. An emissive energy harvester (EEH) is a device that can generate energy from emitting thermal radiation into the clear sky. We calculate how much power is thermodynamically available, using a location in Oklahoma as a case study. We discuss two possible ways to make such a device: A thermal EEH (analogous to solar thermal power generation) and an optoelectronic EEH (analogous to photovoltaic power generation). For the latter, we propose using a rectifying antenna, and we discuss its operating principles, efficiency limits, system design considerations, and possible technological implementations.
- 2 September 2013
- Published by AIP Publishing in Applied Physics Letters
- Vol. 103 (10) , 101104
Enhancement of absorption and color contrast in ultra-thin highly absorbing optical coatings
Recently a new class of optical interference coatings was introduced which comprises ultra-thin, highly absorbing dielectric layers on metal substrates. We show that these lossy coatings can be augmented by an additional transparent subwavelength layer. We fabricated a sample comprising a gold substrate, an ultra-thin film of germanium with a thickness gradient, and several alumina films. The experimental reflectivity spectra showed that the additional alumina layer increases the color range that can be obtained, in agreement with calculations. More generally, this transparent layer can be used to enhance optical absorption, protect against erosion, or as a transparent electrode for optoelectronic devices.
- 2 September 2013
- Published by AIP Publishing in Applied Physics Letters
- Vol. 103 (10) , 101104
Enhancement of absorption and color contrast in ultra-thin highly absorbing optical coatings
Recently a new class of optical interference coatings was introduced which comprises ultra-thin, highly absorbing dielectric layers on metal substrates. We show that these lossy coatings can be augmented by an additional transparent subwavelength layer. We fabricated a sample comprising a gold substrate, an ultra-thin film of germanium with a thickness gradient, and several alumina films. The experimental reflectivity spectra showed that the additional alumina layer increases the color range that can be obtained, in agreement with calculations. More generally, this transparent layer can be used to enhance optical absorption, protect against erosion, or as a transparent electrode for optoelectronic devices.
- 23 July 2012
- Published by American Chemical Society (ACS) in Nano Letters
- Vol. 12 (8) , 4300-4304
Screening-Engineered Field-Effect Solar Cells
Photovoltaics (PV) are a promising source of clean renewable energy, but current technologies face a cost-to-efficiency trade-off that has slowed widespread implementation.(1, 2) We have developed a PV architecture—screening-engineered field-effect photovoltaics (SFPV)—that in principle enables fabrication of low-cost, high efficiency PV from virtually any semiconductor, including the promising but hard-to-dope metal oxides, sulfides, and phosphides.(3) Prototype SFPV devices have been constructed and are found to operate successfully in accord with model predictions.
- 23 July 2012
- Published by American Chemical Society (ACS) in Nano Letters
- Vol. 12 (8) , 4300-4304
Screening-Engineered Field-Effect Solar Cells
Photovoltaics (PV) are a promising source of clean renewable energy, but current technologies face a cost-to-efficiency trade-off that has slowed widespread implementation.(1, 2) We have developed a PV architecture—screening-engineered field-effect photovoltaics (SFPV)—that in principle enables fabrication of low-cost, high efficiency PV from virtually any semiconductor, including the promising but hard-to-dope metal oxides, sulfides, and phosphides.(3) Prototype SFPV devices have been constructed and are found to operate successfully in accord with model predictions.
- 1 November 2011
- Published by Elsevier in Chemical Physics Letters
- Vol. 516 (4-6) , 115-124
Ambiguities in surface nonlinear spectroscopy calculations
- 1 November 2011
- Published by Elsevier in Chemical Physics Letters
- Vol. 516 (4-6) , 115-124
Ambiguities in surface nonlinear spectroscopy calculations