We study the problem of change-point detection and localisation for functional data sequentially observed on a general $d$-dimensional space, where we allow the functional curves to be either sparsely or densely sampled. Data of this form naturally arise in a wide range of applications such as biology, neuroscience, climatology and finance. To achieve such a task, we propose a kernel-based algorithm named functional seeded binary segmentation (FSBS). FSBS is computationally efficient, can handle discretely observed functional data, and is theoretically sound for heavy-tailed and temporally-dependent observations. Moreover, FSBS works for a general $d$-dimensional domain, which is the first in the literature of change-point estimation for functional data. We show the consistency of FSBS for multiple change-point estimation and further provide a sharp localisation error rate, which reveals an interesting phase transition phenomenon depending on the number of functional curves observed and the sampling frequency for each curve. Extensive numerical experiments illustrate the effectiveness of FSBS and its advantage over existing methods in the literature under various settings. A real data application is further conducted, where FSBS localises change-points of sea surface temperature patterns in the south Pacific attributed to El Ni\~{n}o.

To handle complex instances, we revisit a divide-and-conquer approach to break down the complexity: instead of few complex BaB calls, we rely on many small {\em partial} MILP calls. The crucial step is to select very few but very important ReLUs to treat using (costly) binary variables. The previous attempts were suboptimal in that respect. To select these important ReLU variables, we propose a novel {\em solution-aware} ReLU scoring ({\sf SAS}), as well as adapt the BaB-SR and BaB-FSB branching functions as {\em global} ReLU scoring ({\sf GS}) functions. We compare them theoretically as well as experimentally, and {\sf SAS} is more efficient at selecting a set of variables to open using binary variables. Compared with previous attempts, SAS reduces the number of binary variables by around 6 times, while maintaining the same level of accuracy. Implemented in {\em Hybrid MILP}, calling first $α,β$-CROWN with a short time-out to solve easier instances, and then partial MILP, produces a very accurate yet efficient verifier, reducing by up to $40\%$ the number of undecided instances to low levels ($8-15\%$), while keeping a reasonable runtime ($46s-417s$ on average per instance), even for fairly large CNNs with 2 million parameters.

We study the problem of change-point detection and localisation for functional data sequentially observed on a general d -dimensional space, where we allow the functional curves to be either sparsely or densely sampled. Data of this form naturally arise in a wide range of applications such as biology, neuroscience, climatology and finance. To achieve such a task, we propose a kernel-based algorithm named functional seeded binary segmentation (FSBS). FSBS is computationally efficient, can handle discretely observed functional data, and is theoretically sound for heavy-tailed and temporally-dependent observations. Moreover, FSBS works for a general d -dimensional domain, which is the first in the literature of change-point estimation for functional data.

Reinforcement learning from human feedback (RLHF) is fundamentally limited by the capacity of humans to correctly evaluate model output. To improve human evaluation ability and overcome that limitation this work trains "critic" models that help humans to more accurately evaluate model-written code. These critics are themselves LLMs trained with RLHF to write natural language feedback highlighting problems in code from real-world assistant tasks. On code containing naturally occurring LLM errors model-written critiques are preferred over human critiques in 63% of cases, and human evaluation finds that models catch more bugs than human contractors paid for code review. We further confirm that our fine-tuned LLM critics can successfully identify hundreds of errors in ChatGPT training data rated as "flawless", even though the majority of those tasks are non-code tasks and thus out-of-distribution for the critic model. Critics can have limitations of their own, including hallucinated bugs that could mislead humans into making mistakes they might have otherwise avoided, but human-machine teams of critics and contractors catch similar numbers of bugs to LLM critics while hallucinating less than LLMs alone.

According to the ONS (Office for National Statistics) around 1.5 million jobs in England alone could be at risk from automation. Deloitte finds that over the next two decades, up to 35% of all jobs could be at high risk of being replaced by automated systems. If the rise of automation has the potential to impact millions of jobs, will workers want to adapt and learn new skills if their jobs are at risk? A two-year Commission on Workers and Technology is currently assessing the impact automation and artificial intelligence (AI) could have on the UK's workforce. Yvette Cooper MP, who is chair of the Commission, wrote in The Guardian: "Trades unions and communities can't just stand by and hope for the best. If we want technological change to benefit everyone rather than widening inequality, then we need to start preparing now."

Replacing bank and insurance workers with machines risks creating a dependency on outside technology companies beyond the reach of regulators, the global Financial Stability Board (FSB) said on Wednesday. The FSB, which coordinates financial regulation across the Group of 20 Economies (G20), said in its first report on artificial intelligence (AI) and machine learning that the risks they pose need monitoring. AI and machine learning refer to technology that is replacing traditional methods to assess the creditworthiness of customers, to crunch data, price insurance contracts and spot profitable trades across markets. There are no international regulatory standards for AI and machine learning, but the FSB left open whether new rules are needed. Data on rapidly growing usage of AI is largely unavailable, leaving regulators unsure about the impact of potentially new and unexpected links between markets and banks, the report said.

Replacing bank and insurance workers with machines risks creating a dependency on outside technology companies beyond the reach of regulators, the global Financial Stability Board (FSB) said on Wednesday. The FSB, which coordinates financial regulation across the Group of 20 Economies (G20), said in its first report on artificial intelligence (AI) and machine learning that the risks they pose need monitoring. AI and machine learning refer to technology that is replacing traditional methods to assess the creditworthiness of customers, to crunch data, price insurance contracts and spot profitable trades across markets. There are no international regulatory standards for AI and machine learning, but the FSB left open whether new rules are needed. Data on rapidly growing usage of AI is largely unavailable, leaving regulators unsure about the impact of potentially new and unexpected links between markets and banks, the report said.

Banks and hedge funds that rely on artificial intelligence threaten to inject risks into the financial system that could exacerbate a future crisis, according to global regulators. The financial industry's rush to adopt AI raises the potential that firms will become overly dependent on technologies that herd them toward the same view of risks and could "amplify financial shocks," according to a study published on Wednesday by the Financial Stability Board, a panel of regulators that includes the U.S. Federal Reserve and European Central Bank. "AI and machine learning applications show substantial promise if their specific risks are properly managed," the FSB said in a report that called for additional monitoring and testing of robotic technologies designed to lessen human involvement. "Taken as a group, universal banks' vulnerability to systemic shocks may grow if they increasingly depend on similar algorithms or data streams." The FSB, headed by Bank of England Governor Mark Carney, said that many of the technologies are being designed and tested in a period of low volatility in financial markets, and, as a result, "may not suggest optimal actions in a significant economic downturn or in a financial crisis."