This is the edition of 9 November 2026. See the latest news.
News
Updated three times a day
IBM Research·9 November 2026
IBM Research describes the Praxis Policy Engine, which it presents as a programmable security layer for AI agents. According to IBM, the engine applies policy controls across the AI stack, with coverage that spans inference, gateways and the agents themselves.
Read the original at IBM Research ↗
ETH Zurich·9 November 2026
ETH Zurich reports work on microbots that are driven by magnetic fields and generate flow inside blood vessels. The work addresses clots that restrict circulation so severely that clot-dissolving drugs can hardly reach them. The researchers propose the flow these robots create as a possible way around that obstacle.
Read the original at ETH Zurich ↗
MIT News·9 November 2026
MIT and New York University researchers introduce FLEX, a feature-editing pipeline that lets people steer robot motion planning through conversational text. Using a large language model, the robot is meant to infer the broader intent behind an instruction, such as avoiding a laptop while carrying coffee. The team reports tests in simulated scenarios and will present the study at the Conference on Robot Learning.
Read the original at MIT News ↗
New research papers
arXiv
Preprints from the arXiv categories Robotics (cs.RO) and Artificial Intelligence (cs.AI). Preprints are not peer reviewed.
01
arXiv · cs.RO6 November 2026
The authors examine what happens when sampling-based planners such as MPPI search hard against reward models learned from human preferences. They report that stronger search can cut true return by up to 76% while held-out label accuracy stays high, and they derive a regularized temperature rule that they find safest among tested fixes.
Read the original at arXiv ↗
02
arXiv · cs.RO6 November 2026
The authors test whether OpenVLA obeys negated instructions on LIBERO-Goal and probe where it encodes its executed goal using patching, probes and steering. They report that negation has no detectable effect, with the forbidden task executed in 79% of rollouts, and that goal edits only work early in the network or across every action token.
Read the original at arXiv ↗
03
arXiv · cs.AI6 November 2026
The authors build ABHBench, a multimodal benchmark drawn from the game Detroit: Become Human, to assess how foundation models choose when acting as android agents in moral dilemmas, scoring harmlessness, obedience and self-preservation. Across 12 models, they report that 75% shift significantly between text-only and situated multimodal versions.
Read the original at arXiv ↗
04
arXiv · cs.RO6 November 2026
The authors propose a counterfactual audit for grasp-feasibility predictors that places an obstacle in the approach corridor and has a motion planner relabel the grasp as infeasible. Across eight predictors, they report pair AUROC falling to chance, and note that a clearance-aware corridor reading recovers performance only for trained obstacle placements.
Read the original at arXiv ↗
05
arXiv · cs.AI6 November 2026
The authors frame the decision to verify or steer an agent mid-trajectory as a causal inference problem and build a framework for estimating dynamic effects of randomized interventions from a modest set of trajectories. They describe using the estimated parameters to build safety leaderboards for audit prioritization, applying the method to tau-bench.
Read the original at arXiv ↗
06
arXiv · cs.RO6 November 2026
The authors run a controlled simulation study of coding agents that write and refine robot policies, varying model capability, skill priors, feedback and curriculum across nine manipulation tasks and 28K rollouts. They report that the main bottleneck tends to be turning available information into correct policy updates rather than accessing that information.
Read the original at arXiv ↗
07
arXiv · cs.AI6 November 2026
The authors survey 289 U.S. adults on 45 scenarios across nine life domains to learn which tasks people would hand to AI agents. They report that trust, desired AI leadership and expected savings predict greater automation preference, while concern about lost personal fulfillment predicts less, with respondents favoring retained human oversight.
Read the original at arXiv ↗
08
arXiv · cs.RO6 November 2026
The authors introduce STRIDE, which trains humanoid walking with a reward based on following commanded accelerations measured by the onboard IMU, combined with off-policy learning, high joint gains and symmetry augmentation. In simulation, they report Unitree G1 and Booster T1 walking within 14 minutes of samples and 81% fewer training falls.
Read the original at arXiv ↗
09
arXiv · cs.RO6 November 2026
The authors report that AMP1, a harness letting a generalist AI model directly command single-arm, bimanual, and mobile manipulators, scores 85.6% zero-shot on RoboLab-120 versus 49.2% and 42.9% for top VLA and WAM policies, without demonstrations or task training. They also describe 104 successes in 108 real trials and say adding cross-episode memory cut RoboDojo bimanual failures 4.6x.
Read the original at arXiv ↗
AI-generated summaries. Items, titles and links come from the publishers' public news feeds and from arXiv, and each title links to the original. Summaries are written and checked automatically from the source text and are not individually reviewed by a person; the original is authoritative. Research papers are preprints, not peer reviewed. Llull Lab is independent of the publishers listed. How it is made.