Alexander Rakhlin named director of the MIT Statistics and Data Science Center
An expert in machine learning, statistics, and computation, Rakhlin succeeds Professor Ankur Moitra.
More from the blog
MIT Quantum Initiative launches postdoctoral fellowship program
The Institute welcomes its first cohort of QMIT Fellows this fall to advance interdisciplinary quantum research.
TimesFM-3: A zero-shot foundation model for multivariate forecasting
Data Management
How an MIT research project became a global programming language
With millions of users across the world, Julia has been used to conduct cutting-edge research and to design new drugs, jet engines, heat pumps, and more.
AI helps design new materials that work in the real world
The “CrysVCD” tool developed at MIT could cut the huge amounts of time and money spent on screening out chemically unstable designs.
Related papers
Constitutional Midtraining: Content Presence Drives Alignment Gains
Post-training alignment is often shallow, eroding under fine-tuning. It remains untested as to whether constitutional midtraining interventions can produce durable alignment when cleanly isolated from post-training. We build a 394M-token constitutional corpus from Anthropic's Constitution and apply constitutional midtraining at 120B scale, where principled, values-based content is inserted into midtraining. A 2x2 design (curriculum ordering x deliberative reasoning) was used to produce four constitutionally midtrained conditions, plus a control, which were evaluated on self-generated and established benchmarks including alignment under pressure, value conflict resolution, blackmail, and emergent misalignment. All models were evaluated across three stages: post-midtraining, post-SFT, and post-benign fine-tuning. Constitutionally midtrained models outperformed the control on alignment generalization and durability, notably on blackmail: SFT instilled a blackmail propensity in all models, but constitutional midtraining blunted it, with the advantage surviving benign fine-tuning (-17.5pp). This durability did not extend to settings that required active resistance to in-context pressure or conflict, where the advantage attenuates after SFT. The presence of constitutional content at midtraining also mattered more than its structure, and constitutional midtraining incurred no capability cost, on average, at any stage (MMLU, ARC-Easy, piqa, GSM8K). A modest amount of constitutional content at midtraining could therefore yield broad, persistent alignment gains, offering a cheap, complementary addition to SFT-centered pipelines. Code, data, and models are available.
Asynchronous Federated Reinforcement Learning for Adaptive Resource Slicing and Low-Latency Task Offloading in Heterogeneous 6G Edge Computing Networks
The emerging paradigm of 6G wireless communication networks envisions ultra-reliable low-latency communication (URLLC), massive machine-type communications (mMTC), and pervasive edge computing intelligence. In heterogeneous mobile edge computing (MEC) networks, dynamically offloading compute-intensive tasks (e.g., augmented reality rendering, connected vehicular telemetry, autonomous robotic control) while orchestrating multi-tenant network slicing under time-varying channel conditions is an NP-hard stochastic optimization problem. Centralized reinforcement learning algorithms suffer from extreme communication overhead, severe backhaul congestion, and severe privacy vulnerabilities. Conversely, standard synchronous Federated Learning (FL) methods encounter severe 'straggler effects' caused by heterogeneous edge device processing capabilities. In this paper, we propose AF-EdgeRL, a novel Byzantine-resilient Asynchronous Federated Reinforcement Learning framework tailored for distributed resource allocation and dynamic task offloading. AF-EdgeRL deploys a distributed Proximal Policy Optimization (PPO) agent across edge servers and end-user devices, combined with a Staleness-Aware Adaptive Weight Aggregator (SAWA) that dynamically adjusts model update gradients based on hardware compute latency and channel state information (CSI). Furthermore, we establish theoretical convergence guarantees under non-convex reinforcement learning objectives. Evaluated on a high-fidelity 6G MEC simulator with real-world mobile mobility traces (Telecom Italia Milano dataset), AF-EdgeRL reduces end-to-end task execution latency by 41.2%, achieves 99.999% URLLC deadline compliance, and decreases edge energy consumption by 32.6% compared to state-of-the-art synchronous FedRL and centralized DRL baselines.
AoI Minimization in Heterogeneous MEC Networks: A Federated Learning-Assisted Hybrid DRL and Convex Approach
This paper investigates a dynamic heterogeneous mobile edge computing network (HMECN), where mobile devices (MDs) could offload their full tasks to a small base station (SBS) directly or the macro base station (MBS) in direct or relay mode. As age of information (AoI) is a comprehensive and accurate metric to capture the freshness of computation results, we formulate a long-term weighted sum AoI (LWSA) minimization problem in the HMECN by jointly optimizing the offloading decisions of MDs as well as the bandwidth and computation resource allocation of all base stations, subject to energy, delay and peak AoI constraints. To address the formulated non-convex mixed integer nonlinear programming problem, we decompose it into the offloading decision optimization (ODO) top-problem and the resource allocation optimization (RAO) sub-problem. Based on the decomposition, we propose a federated learning (FL)-assisted hybrid DRL and convex approach that is comprised of a safe multi-agent DRL algorithm, convex optimization and FL. The ODO top-problem is solved by the safe multi-agent DRL algorithm, which strictly ensures that the actions of each agent do not exceed its energy constraint and then paves the way for using convex optimization to solve the RAO sub-problem. FL is used to alleviate the training instability problem aggravated by multi-agent settings via breaking the limitation of partial knowledge for each individual agent. Simulation results demonstrate the superiority of the proposed approach in terms of the LWSA, convergence, scalability and robustness in dynamic environments.
Fermi-Dirac thermal measurements: A framework for quantum hypothesis testing and semidefinite optimization
Quantum measurements are the means by which we recover messages encoded into quantum states. They are at the forefront of quantum hypothesis testing, wherein the goal is to perform an optimal measurement for arriving at a correct conclusion. Mathematically, a measurement operator is Hermitian with eigenvalues in [0,1]. By noticing that this constraint on each eigenvalue is the same as that imposed on fermions by the Pauli exclusion principle, we interpret every eigenmode of a measurement operator as an independent effective fermionic mode. Under this perspective, various objective functions in quantum hypothesis testing can be viewed as the total expected energy associated with these fermionic occupation numbers. By instead fixing a temperature and minimizing the total expected fermionic free energy, we find that optimal measurements for these modified objective functions are Fermi-Dirac thermal measurements, wherein their eigenvalues are specified by Fermi-Dirac distributions. In the low-temperature limit, their performance closely approximates that of optimal measurements for quantum hypothesis testing, and we show that their parameters can be learned by classical or hybrid quantum-classical optimization algorithms. This leads to a new quantum machine-learning model, termed Fermi-Dirac machines, consisting of parameterized Fermi-Dirac thermal measurements-an alternative to quantum Boltzmann machines based on thermal states. Beyond hypothesis testing, we show how general semidefinite optimization problems can be solved using this approach, leading to a novel paradigm for semidefinite optimization on quantum computers, in which the goal is to implement thermal measurements rather than prepare thermal states. Finally, we propose quantum algorithms for implementing Fermi-Dirac thermal measurements, and we also propose second-order hybrid quantum-classical optimization algorithms.