Managing extreme AI risks amid rapid progress.

Preparation requires technical research and development, as well as adaptive, proactive governance.

Evaluating Protein Transfer Learning with TAPE.

Machine learning applied to protein sequences is an increasingly popular area of research. Semi-supervised learning for proteins has emerged as an important paradigm due to the high cost of acquiring supervised protein labels, but the current literature is fragmented when it comes to datasets and standardized evaluation techniques. To facilitate progress in this field, we introduce the Tasks Assessing Protein Embeddings (TAPE), a set of five biologically relevant semi-supervised learning tasks spread across different domains of protein biology. We curate tasks into specific training, validation, and test splits to ensure that each task tests biologically relevant generalization that transfers to real-life scenarios. We benchmark a range of approaches to semi-supervised protein representation learning, which span recent work as well as canonical sequence learning techniques. We find that self-supervised pretraining is helpful for almost all models on all tasks, more than doubling performance in some cases. Despite this increase, in several cases features learned by self-supervised pretraining still lag behind features extracted by state-of-the-art non-neural techniques. This gap in performance suggests a huge opportunity for innovative architecture design and improved modeling paradigms that better capture the signal in biological sequences. TAPE will help the machine learning community focus effort on scientifically relevant problems. Toward this end, all data and code used to run these experiments are available at https://github.com/songlab-cal/tape.

Deciphering the role of a coleopteran steering muscle via free flight stimulation.

Testing hypotheses of neuromuscular function during locomotion ideally requires the ability to record cellular responses and to stimulate the cells being investigated to observe downstream behaviors [1]. The inability to stimulate in free flight has been a long-standing hurdle for insect flight studies. The miniaturization of computation and communication technologies has delivered ultra-small, radio-enabled neuromuscular recorders and stimulators for untethered insects [2-8]. Published stimulation targets include the areas in brain potentially responsible for pattern generation in locomotion [5], the nerve chord for abdominal flexion [9], antennal muscles [2, 10], and the flight muscles (or their excitatory junctions) [7, 11-13]. However, neither fine nor graded control of turning has been demonstrated in free flight, and responses to the stimulation vary widely [2, 5, 7, 9]. Technological limitations have precluded hypotheses of function validation requiring exogenous stimulation during flight. We investigated the role of a muscle involved in wing articulation during flight in a coleopteran. We set out to identify muscles whose stimulation produced a graded turning in free flight, a feat that would enable fine steering control not previously demonstrated. We anticipated that gradation might arise either as a function of the phase of muscle firing relative to the wing stroke (as in the classic fly b1 muscle [14, 15] or the dorsal longitudinal and ventral muscles of moth [16]), or due to regulated tonic control, in which phase-independent summation of twitch responses produces varying amounts of force delivered to the wing linkages [15, 17, 18].

A biological micro actuator: graded and closed-loop control of insect leg motion by electrical stimulation of muscles.

In this study, a biological microactuator was demonstrated by closed-loop motion control of the front leg of an insect (Mecynorrhina torquata, beetle) via electrical stimulation of the leg muscles. The three antagonistic pairs of muscle groups in the front leg enabled the actuator to have three degrees of freedom: protraction/retraction, levation/depression, and extension/flexion. We observed that the threshold amplitude (voltage) required to elicit leg motions was approximately 1.0 V; thus, we fixed the stimulation amplitude at 1.5 V to ensure a muscle response. The leg motions were finely graded by alternation of the stimulation frequencies: higher stimulation frequencies elicited larger leg angular displacement. A closed-loop control system was then developed, where the stimulation frequency was the manipulated variable for leg-muscle stimulation (output from the final control element to the leg muscle) and the angular displacement of the leg motion was the system response. This closed-loop control system, with an optimized proportional gain and update time, regulated the leg to set at predetermined angular positions. The average electrical stimulation power consumption per muscle group was 148 µW. These findings related to and demonstrations of the leg motion control offer promise for the future development of a reliable, low-power, biological legged machine (i.e., an insect-machine hybrid legged robot).

Insect-machine hybrid system.

We demonstrated the remote control of insects in free flight via an implantable radio-equipped miniature neural stimulating system. This paper summarizes these results. The pronotum mounted system consisted of neural stimulators, muscular stimulators, a radio transceiver-equipped microcontroller and a microbattery. Flight initiation, cessation and elevation control were accomplished through neural stimulus of the brain which elicited, suppressed or modulated wing oscillation. Turns were triggered through the direct muscular stimulus of either of the basalar muscles. We characterized the response times, success rates, and free-flight trajectories elicited by our neural control systems in remotely-controlled beetles. We believe this type of technology will open the door to in-flight perturbation and recording of insect flight responses.