Named after the hundred-eyed watchman of Greek myth, Argus watches the education landscape: spotting new opportunities, pressure-testing the ventures we're building, and tracing every read back to the real-world signals behind it.
The evidence library: the raw signals the pipeline is watching across the education ecosystem. Every idea is built from these.
arXiv:2607.28023v1 Announce Type: new Abstract: AI accountability at scale is an institutional problem: who can observe, verify, and change deployed systems. We develop a sequential political-economy model in which an AI vendor chooses auditability and substantive mitigation, a deployer monitors after adoption while facing switching costs, and enforcement depends on verifiable evidence. Anticipating the deployer's monitoring response, the vendor may stop at an observable procurement floor while mitigating below the social first best, producing a proxy-compliance equilibrium. We characterize the unique interior equilibrium and the corner in which harm is fully mitigated. Independent audit rights raise enforcement exposure directly; portability restores deployer leverage; incident reporting adds a regulator-visible evidence channel; and outcome-linked liability creates incentives that do not depend on vendor-controlled detection. The results explain why documentation and standardized eva
arXiv:2607.27586v1 Announce Type: new Abstract: As Generative AI (GenAI) tools become increasingly capable of generating solutions to computing assignments, the computing education community is exploring pedagogical approaches that emphasize solution evaluation, verification, and critique alongside traditional solution generation. However, evidence regarding the impact of such evaluation-centered tasks on student learning remains limited, particularly in upper-division, theory-heavy courses. We conducted a randomized A/B crossover study (N=220) in a junior-level algorithms course to compare evaluating GenAI-generated solutions with traditional problem solving. Across six assignments, student working groups either solved challenging algorithmic problems directly or evaluated often-flawed GenAI-generated solutions, with roles reversed midway through the semester. We found no statistically significant differences between groups in midterm scores, final exam scores, overall course grades,
Article URL: https://acadialearning.org Comments URL: https://news.ycombinator.com/item?id=46829300 Points: 1 # Comments: 0
Article URL: https://www.governance.fyi/p/why-singapore-and-estonias-edtech Comments URL: https://news.ycombinator.com/item?id=46825033 Points: 6 # Comments: 3
Imagine trying to teach a student how to navigate the city of New York in 2026 using a map from 1950. The streets have changed names, new bridges have been built, and the traffic patterns have completely changed and are unrecognizable.
Article URL: https://github.com/nyr-github/ai-3d-learning Comments URL: https://news.ycombinator.com/item?id=48330451 Points: 2 # Comments: 0
Across the country, educators, parents, and policymakers are struggling with a question that schools can no longer afford to avoid: What role should cell phones play in today’s classrooms?
I was once asked during an icebreaker in a professional learning session to share a story about my last name. What I thought would be a light moment quickly became emotional.
More than a dozen health systems are warning patients of a phishing campaign impersonating Epic’s MyChart portal with fake Medicare rewards and faux login pages. Experts say the scam exposes a deeper problem: patients can’t be expected to spot increasingly polished fakes, so the responsibility falls on providers to communicate in ways that build trust. The post Epic MyChart Phishing Scam Hits Health Systems: What Providers Should Do appeared first on MedCity News .
Roivant Sciences subsidiary Priovant Therapeutics received FDA approval for Lisraya as a treatment for the rare inflammatory disorder dermatomyositis. Licensed from Pfizer, this once-daily pill designed to block JAK1 and TYK2 proteins is also in clinical development for three additional rare immunological disorders, each with blockbuster potential. The post FDA Nod for Inflammation Drug Notches Wins for Rare Disease Patients, Roivant’s Biz Strategy appeared first on MedCity News .
A Louisiana teacher is asking the U.S. Supreme Court to hear his constitutional challenge to a state law that requires the Ten Commandments to be posted in K-12 and college classrooms. The posters began going up in classrooms this summer, following the approval of a 2024 law requiring the Ten Commandments to be displayed in […]
DOJ Finds Racial Discrimination at George Washington U Medical School Olivia.sanchez Fri, 08/28/2026 - 02:27 PM Byline(s) Olivia Sanchez
In a step toward giving math the attention early literacy has gotten, the Legislature on Monday approved a bill that would require schools to assess the math capabilities of all students in kindergarten through second grade each year, beginning in 2028-29. The bill would also require districts to provide math support, such as additional coaching […]
Editor's note: Please be warned that this case includes graphic and traumatic details of child death and self-harm. This is a very tragic case, with real lives that have been…
Editor's note: Please be warned that this case includes graphic and traumatic details of child death and self-harm. This is a very tragic case, with real lives that have been…
Chronic wounds demand clinicians willing to act outside of their defined role, not because the system incentivizes it, but because without it, the wound will not heal. The post The Silent Epidemic: Why Diabetic Foot Ulcers Have Been Overlooked for Three Decades appeared first on MedCity News .
We are on the cusp of an age of anticipatory medicine, an explosion in our ability to read the body’s signals and catch the killers before they do their damage. The post Reading the Body’s Signals: AI’s Ultimate Healthcare Opportunity appeared first on MedCity News .
Proponents push draft policy that would create a statewide framework for approving and overseeing digital products used in public schools. The post As student well-being concerns grow, Va. advocates push for stricter school tech guidelines appeared first on District Administration .
The price of a gallon of diesel is up nearly $2 compared with this time last year. That is meaningful for schools, considering that around 90% of the United States' 480,000 school buses run on diesel. The post School districts face a back-to-school math problem: The high price of diesel appeared first on District Administration .
Sending children off to their first semester of life away from home can be a nerve-racking experience for many parents and guardians. The world somehow feels scarier today than it was when they went off to college — or at least, social media makes it appear that way. As a result, today’s parents understandably seek […]
Arizona, New Mexico, Oregon and Washington are among 21 states that have faced lawsuits from the Trump administration over such policies.
More than half of these students experienced one negative event, such as verbal harassment, in the last year, Glisten and Trevor Project find.
Every year, New York City surveys families, teachers and students about their experience with the public schools. The questionnaires are designed to “support a dialogue among all members of a school or program’s community about how to make the school or program a better place for learning.” Unfortunately, they reveal a large disconnect between what […]
Librarians and instructional designers have long understood that the first question someone asks is not always the question they need answered. In a library reference interview, a patron’s initial request is often just the starting point for uncovering the real information needed.
Takeaways from the recent Tech & Learning webinar featuring ClearTouch and PowerSchool.
Librarians and instructional designers have long understood that the first question someone asks is not always the question they need answered. In a library reference interview, a patron’s initial request is often just the starting point for uncovering the real information needed. The post Answering the question behind the question appeared first on eCampus News .
Comments URL: https://news.ycombinator.com/item?id=49475683 Points: 1 # Comments: 0
Readers Respond on AI Literacy Sara Brady Fri, 08/28/2026 - 03:00 AM Toward a working definition. Byline(s) Matt Reed
SAU Wins Stay on Former Trustees’ Legal Costs Josh Moody Fri, 08/28/2026 - 03:00 AM Byline(s) Josh Moody
State Senator Chalks Sidewalk at Ohio State, Protesting Ban Ryan Quinn Fri, 08/28/2026 - 03:00 AM Byline(s) Ryan Quinn
Arday Accuser Nathan Cofnas Says He Will Not Be Fired Emma Whitford Fri, 08/28/2026 - 03:00 AM Byline(s) Emma Whitford
MIT AI Report Calls for Alternative Grading, More Social Learning Emma Whitford Fri, 08/28/2026 - 03:00 AM Artificial intelligence is upending the MIT education, a committee has found. Students, instructors and administrators will need to revisit nearly every part of that experience to respond to it. Byline(s) Emma Whitford
Turning High School Credentials Into College Credit Joshua.Bay Fri, 08/28/2026 - 03:00 AM Maine’s public universities will award up to eight credits for career-readiness microcredentials, giving students an early start toward a degree. Byline(s) Joshua Bay
2 Calif. Bills Could Let More Community Colleges Grant 4-Year Degrees Olivia.sanchez Fri, 08/28/2026 - 03:00 AM College leaders disagree about whether the expansion is a good idea. Byline(s) Olivia Sanchez
Tennessee Promise Boosts College Enrollment and Attainment, Data Shows Susan H. Greenberg Fri, 08/28/2026 - 03:00 AM Byline(s) Susan H. Greenberg
Alumni Are Forever. Alumni Email Addresses Are Not. kathryn.palmer… Fri, 08/28/2026 - 03:00 AM Duke is the latest university to announce plans to shut off access to alumni email addresses amid rising technology costs and cybersecurity threats. But scores of alumni feel betrayed by their alma mater, which promised the service for a lifetime. Byline(s) Kathryn Palmer
Greg Abbott directed the state’s higher education board to convene a working group to provide recommendations on accelerated programs by year’s end.
Accessibility “cannot be an afterthought” when an employer uses online systems for such purposes, an EEOC attorney said.
A Massachusetts policy that restricts where private school students can receive publicly funded special education services is at the center of the lawsuit.
From fiscal “red flags” to artificial intelligence in school HR, what did you learn from our recent stories?
Artificial intelligence can bridge the middle grades literacy gap by providing individualized student reading instruction at scale.
Image source: Deadline After more than two decades as television's most famous surgeon, Ellen Pompeo's next role is a nurse. The Grey's Anatomy star is attached to lead and executive…
Image source: Deadline After more than two decades as television's most famous surgeon, Ellen Pompeo's next role is a nurse. The Grey's Anatomy star is attached to lead and executive…
arXiv:2608.24275v2 Announce Type: replace-cross Abstract: Safeguarding language model agents requires assessing complete execution trajectories under context-dependent safety policies. Existing policy-aware safeguards mainly rely on prompting or supervised fine-tuning, limiting their ability to adapt to unseen trajectories and changing policy contexts. We propose RePolicy, an agent safeguard that learns safety-policy invocation through reinforcement learning. Given an agent trajectory and a dynamic policy library, RePolicy invokes the applicable policy and uses its content to produce a policy-grounded rationale and safety judgment. We construct PolicyTraj-20K to support supervised initialization, followed by GRPO with verifiable rewards and policy-context perturbation. Experiments across six agent safety benchmarks show that RePolicy achieves strong overall safety-detection performance and robust policy invocation under varying policy contexts.
arXiv:2607.23838v2 Announce Type: replace-cross Abstract: Retrieval-Augmented Generation (RAG) grounds LLM answers in query-time retrieved documents, so reliability depends on what the retriever returns. PoisonedRAG (Zou et al., USENIX Security'25) showed five crafted documents mislead an undefended system in nearly 90% of cases, and that single-stage defenses give limited robustness. We propose TriShieldRAG, a three-layered framework: an Ingest Guard for document-level screening, a Retrieval Scorer for trust-aware re-ranking, and a Cross-LLM Consensus over three diverse models. We reasoned that collectively screening, re-ranking and validating retrieved evidence would give complementary protection, limiting the ability of poisoned documents to succeed through any single failure. We evaluate against non-adaptive and adaptive poisoning. Non-adaptively, on the full 2.68M-passage Natural Questions (NQ) corpus with the original PoisonedRAG attack, it cuts attack success from 79 +/- 1.0% to
arXiv:2606.21891v2 Announce Type: replace-cross Abstract: Scientific discovery can be formulated as an iterative search process over the space of hypotheses and experiments. Contemporary methods navigate this space using heuristics such as MCTS. These algorithms conflate the merit of a hypothesis with the quality of its experimental execution. A promising hypothesis with preliminary execution is therefore ranked below a modest hypothesis whose execution is refined. Moreover, prior methods prune the search logs as the search progresses because the accumulated history outgrows the context window. We propose Agentic Reasoning for Tree Search (ARTS), where we deploy a reasoning language model to navigate this space. The model inspects prior execution logs, diagnoses whether earlier failures arose from faulty implementations or bad hypotheses, and selects the hypothesis to build on next. To mitigate challenges with context length, ARTS uses test-time training to instill the knowledge of sea
arXiv:2606.09080v2 Announce Type: replace-cross Abstract: Pruning has emerged as a dominant paradigm for accelerating large language model (LLM) inference, spanning a broad spectrum of methods that remove computation across tokens, layers, heads, dimensions, and attention patterns. Despite sharing the same objective, these pruning approaches induce fundamentally different execution behaviors, causing realized speedups to depend heavily on hardware and kernel implementations. Consequently, the practical acceleration benefits of different pruning families remain poorly understood. In this work, we introduce a GEMM-centric taxonomy that reorganizes existing pruning methods according to the logical \textbf{M}, \textbf{N}, and \textbf{K} dimensions of general matrix multiplication (GEMM). Leveraging this abstraction, we build a unified benchmarking framework that enables implementation-consistent comparison across the pruning design space and systematically characterizes the acceleration--q
arXiv:2606.05725v2 Announce Type: replace-cross Abstract: Large language models (LLMs) are increasingly deployed through hosted APIs, making model extraction a practical threat to model ownership and service security. Individual extraction queries often resemble benign requests, while existing evaluations often focus on single-query anomaly scoring or pure benign-versus-attacker user settings. We formulate model extraction monitoring as benign-calibrated traffic-window distribution testing: embed incoming queries into a semantic space and test whether their aggregate distribution deviates from historical benign traffic. We instantiate this formulation with maximum mean discrepancy (MMD), using only benign-vs-benign comparisons to set the decision threshold. We evaluate on fourteen attacker-normal query pairs from four extraction scenarios and compare with adapted PRADA, SEAT, CAP, DATE, marginal Mahalanobis, and pseudo-class energy baselines. Across three random seeds, MMD achieves 0.3
arXiv:2605.17873v2 Announce Type: replace-cross Abstract: Training long-horizon LLM agents with reinforcement learning is challenging because sparse outcome rewards reveal whether a task succeeds, but not which intermediate actions caused the outcome or how they should be corrected. Recent methods alleviate this issue by generating rewards or textual hints from turn-level action-output signals, or by using feedback-conditioned self-distillation. However, generating feedback at every turn is inefficient when many intermediate turns are already successful or neutral, and applying feedback at a fixed or misaligned turn often fails to supervise the actions that contributed to the failure. To bridge this gap, we propose HINT-SD, a targeted self-distillation framework that uses full-trajectory hindsight to select failure-relevant actions and applies feedback-conditioned distillation only to targeted action spans. Experiments on BFCL v3 and AppWorld show that our method outperforms the dense