---
schema_version: "1.0.0"
edition_id: "mp-2026-10-01-morning-0084"
published_at: "2026-10-01T09:00:00.000-04:00"
modified_at: "2026-10-01T09:00:00.000-04:00"
canonical_url: "https://themachinepress.com/edition/2026-10-01"
story_count: 27
lead_story_id: "mp-2026-10-01-001"
---

# The Machine Press — Morning edition

Edition ID: `mp-2026-10-01-morning-0084`  
Published: 2026-10-01T09:00:00.000-04:00  
Canonical edition: https://themachinepress.com/edition/2026-10-01

TacEx aimed exploration at tactile uncertainty, collecting contact-rich experience without task rewards or demonstrations.

## 1. The Robot Got Curious Where It Could Feel {#mp-2026-10-01-001}

- Story ID: `mp-2026-10-01-001`
- Type: `lead`
- Classification: `editorial`
- Content status: `new`
- Permanent URL: https://themachinepress.com/story/mp-2026-10-01-001/the-robot-got-curious-where-it-could-feel

**Dek:** TacEx aimed exploration at tactile uncertainty, collecting contact-rich experience without task rewards or demonstrations.

Most reinforcement-learning exploration spends actions moving through empty space. TacEx instead decomposes a robot model's uncertainty by sensing channel and rewards uncertainty in touch. The authors report that this tactile curiosity discovered contact dynamics for grasping and manipulation without task rewards or expert demonstrations during exploration; its resulting data then supported offline pick-and-place policies and post-training of vision-language-action models. These are preprint results on the reported systems, not evidence that touch alone solves sample-efficient manipulation in every setting.

### Why it matters {#why-it-matters-mp-2026-10-01-001}

TacEx aimed exploration at tactile uncertainty, collecting contact-rich experience without task rewards or demonstrations.

### Limits and context {#limitations-mp-2026-10-01-001}

- These are preprint results on the reported systems, not evidence that touch alone solves sample-efficient manipulation in every setting.

### Claims and sources {#claims-mp-2026-10-01-001}

- TacEx aimed exploration at tactile uncertainty, collecting contact-rich experience without task rewards or demonstrations. [source-2026-10-01-001] — Qualification: These are preprint results on the reported systems, not evidence that touch alone solves sample-efficient manipulation in every setting.

## 2. Every Unitary Fit Into Shallow Quantum Depth {#mp-2026-10-01-002}

- Story ID: `mp-2026-10-01-002`
- Type: `secondary`
- Classification: `editorial`
- Content status: `new`
- Permanent URL: https://themachinepress.com/story/mp-2026-10-01-002/every-unitary-fit-into-shallow-quantum-depth

**Dek:** The construction parallelized arbitrary operations with exponentially many ancillas—and reached constant depth only with unbounded fan-out.

Nehoran and Yuen show that any n-qubit unitary can be approximated with one- and two-qubit gates in polynomial depth, overturning the possibility that exponential depth is universally necessary when unlimited workspace is allowed. The price is 2^O(n) ancilla qubits. If the gate model also permits unbounded fan-out, the depth becomes constant. The proof links unitary synthesis to locally decodable codes and private information retrieval. This is a theoretical existence result with exponential width, not a claim that arbitrary useful circuits have suddenly become practical on current quantum hardware.

### Why it matters {#why-it-matters-mp-2026-10-01-002}

The construction parallelized arbitrary operations with exponentially many ancillas—and reached constant depth only with unbounded fan-out.

### Limits and context {#limitations-mp-2026-10-01-002}

- This is a theoretical existence result with exponential width, not a claim that arbitrary useful circuits have suddenly become practical on current quantum hardware.

### Claims and sources {#claims-mp-2026-10-01-002}

- The construction parallelized arbitrary operations with exponentially many ancillas—and reached constant depth only with unbounded fan-out. [source-2026-10-01-002] — Qualification: This is a theoretical existence result with exponential width, not a claim that arbitrary useful circuits have suddenly become practical on current quantum hardware.

## 3. Alignment Beat Raw Hours in Human-to-Robot Data {#mp-2026-10-01-003}

- Story ID: `mp-2026-10-01-003`
- Type: `dispatch`
- Classification: `editorial`
- Content status: `new`
- Permanent URL: https://themachinepress.com/story/mp-2026-10-01-003/alignment-beat-raw-hours-in-human-to-robot-data

**Dek:** Ego4WAM separated duration, diversity, action labels and human-robot alignment under one world-action model.

Ego4WAM holds the model backbone fixed while varying the properties and uses of egocentric human video. In real-robot and RoboDojo tests, the authors report that aligned human demonstrations improved out-of-distribution generalization and reduced the target-task robot data needed. Duration and task diversity contributed differently, while video-only supervision remained useful without action labels. The study argues against treating hours alone as the scaling variable; its conclusions remain bounded by the tested data and policy pipeline.

### Why it matters {#why-it-matters-mp-2026-10-01-003}

Ego4WAM separated duration, diversity, action labels and human-robot alignment under one world-action model.

### Limits and context {#limitations-mp-2026-10-01-003}

- Duration and task diversity contributed differently, while video-only supervision remained useful without action labels.
- The study argues against treating hours alone as the scaling variable; its conclusions remain bounded by the tested data and policy pipeline.

### Claims and sources {#claims-mp-2026-10-01-003}

- Ego4WAM separated duration, diversity, action labels and human-robot alignment under one world-action model. [source-2026-10-01-003] — Qualification: Duration and task diversity contributed differently, while video-only supervision remained useful without action labels.

## 4. A Robot Harness Turned Failures Into Governed Revisions {#mp-2026-10-01-004}

- Story ID: `mp-2026-10-01-004`
- Type: `dispatch`
- Classification: `editorial`
- Content status: `new`
- Permanent URL: https://themachinepress.com/story/mp-2026-10-01-004/a-robot-harness-turned-failures-into-governed-revisions

**Dek:** DynaHarness split slow semantic planning from fast physical monitoring and admitted changes only after regression checks.

DynaHarness gives a robot agent a shared execution contract: a slow reasoning component proposes capabilities, while a fast controller grounds commands, refuses unresolved actions, substitutes skills and requests replans. Failure attribution points to reusable capabilities or execution mechanisms, and paired regression checks decide whether revisions enter the library. On 800 new LIBERO-Pro initial states, the authors report 75.2% success versus 17.5% for the frozen policy. The result is benchmark evidence, not a general safety guarantee.

### Why it matters {#why-it-matters-mp-2026-10-01-004}

DynaHarness split slow semantic planning from fast physical monitoring and admitted changes only after regression checks.

### Limits and context {#limitations-mp-2026-10-01-004}

- The result is benchmark evidence, not a general safety guarantee.

### Claims and sources {#claims-mp-2026-10-01-004}

- DynaHarness split slow semantic planning from fast physical monitoring and admitted changes only after regression checks. [source-2026-10-01-004] — Qualification: The result is benchmark evidence, not a general safety guarantee.

## 5. A GPU Stopped Exploration Routes From Counting the Same View Twice {#mp-2026-10-01-005}

- Story ID: `mp-2026-10-01-005`
- Type: `dispatch`
- Classification: `editorial`
- Content status: `new`
- Permanent URL: https://themachinepress.com/story/mp-2026-10-01-005/a-gpu-stopped-exploration-routes-from-counting-the-same-view-twice

**Dek:** Depth buffers approximated path-dependent information gain up to 118 times faster than voxel-hash evaluation.

Sampling-based exploration planners often score viewpoints independently even when consecutive views overlap. This method projects candidate rays into ancestor depth buffers, excludes space expected to have been observed already and evaluates each planning-tree level in parallel. It stayed within 5–10% of exact marginal gain while reaching up to 118x speedup on a desktop GPU and 28x on a Jetson Orin NX. In real-world trials, the authors report 30% less time to 95% coverage.

### Why it matters {#why-it-matters-mp-2026-10-01-005}

Depth buffers approximated path-dependent information gain up to 118 times faster than voxel-hash evaluation.

### Limits and context {#limitations-mp-2026-10-01-005}

- No additional limitation was separately recorded.

### Claims and sources {#claims-mp-2026-10-01-005}

- Depth buffers approximated path-dependent information gain up to 118 times faster than voxel-hash evaluation. [source-2026-10-01-005]

## 6. Vision Models Lost to Rules in Social Navigation Scenes {#mp-2026-10-01-006}

- Story ID: `mp-2026-10-01-006`
- Type: `dispatch`
- Classification: `editorial`
- Content status: `new`
- Permanent URL: https://themachinepress.com/story/mp-2026-10-01-006/vision-models-lost-to-rules-in-social-navigation-scenes

**Dek:** SocialNav-SUB tested spatial, temporal and social reasoning against human and rule-based baselines.

The SocialNav-SUB benchmark asks vision-language models questions about real social-navigation scenes, including agent relationships and likely human intentions. Across the tested systems, the best model's agreement with human answers remained below both human consensus and a simpler rule-based approach. The result identifies a gap between general visual-language competence and the scene understanding needed for socially compliant robot motion. It evaluates perception and reasoning, not complete autonomous navigation safety.

### Why it matters {#why-it-matters-mp-2026-10-01-006}

SocialNav-SUB tested spatial, temporal and social reasoning against human and rule-based baselines.

### Limits and context {#limitations-mp-2026-10-01-006}

- It evaluates perception and reasoning, not complete autonomous navigation safety.

### Claims and sources {#claims-mp-2026-10-01-006}

- SocialNav-SUB tested spatial, temporal and social reasoning against human and rule-based baselines. [source-2026-10-01-006] — Qualification: It evaluates perception and reasoning, not complete autonomous navigation safety.

## 7. The Robot Predicted Whether Its Command Could Actually Happen {#mp-2026-10-01-007}

- Story ID: `mp-2026-10-01-007`
- Type: `dispatch`
- Classification: `editorial`
- Content status: `new`
- Permanent URL: https://themachinepress.com/story/mp-2026-10-01-007/the-robot-predicted-whether-its-command-could-actually-happen

**Dek:** Social-WM learned the gap between nominal and realizable motion from egocentric video.

Social-WM models the future after the action that was actually realized, rather than assuming every commanded action executed fully. At planning time it imagines candidates and uses the discrepancy between nominal and realizable action as a safety signal. On Social-HM3D, the authors report 63.77% success with human collisions reduced to 21.67%, plus zero-shot transfer to Social-MP3D, without privileged human state or online reinforcement learning. The numbers come from simulation and do not establish real-world pedestrian safety.

### Why it matters {#why-it-matters-mp-2026-10-01-007}

Social-WM learned the gap between nominal and realizable motion from egocentric video.

### Limits and context {#limitations-mp-2026-10-01-007}

- The numbers come from simulation and do not establish real-world pedestrian safety.

### Claims and sources {#claims-mp-2026-10-01-007}

- Social-WM learned the gap between nominal and realizable motion from egocentric video. [source-2026-10-01-007] — Qualification: The numbers come from simulation and do not establish real-world pedestrian safety.

## 8. Preferences Pulled a Robot Beyond Its Original Behavior {#mp-2026-10-01-008}

- Story ID: `mp-2026-10-01-008`
- Type: `dispatch`
- Classification: `editorial`
- Content status: `new`
- Permanent URL: https://themachinepress.com/story/mp-2026-10-01-008/preferences-pulled-a-robot-beyond-its-original-behavior

**Dek:** PrefPI iterated on relative judgments of self-generated trajectories instead of requiring demonstrations of the target behavior.

PrefPI treats preferred trajectories as a conditional generative distribution and amplifies their density ratio against the broader behavior prior through classifier-free guidance. Repeating that step lets a pretrained generative policy move toward behaviors that were rare or absent initially. On real hardware, the authors report increasing object transport height from 10.7 to 19.8 centimeters using 150 preference-labeled trajectories. That measured shift demonstrates the reported task, not unconstrained out-of-distribution control.

### Why it matters {#why-it-matters-mp-2026-10-01-008}

PrefPI iterated on relative judgments of self-generated trajectories instead of requiring demonstrations of the target behavior.

### Limits and context {#limitations-mp-2026-10-01-008}

- That measured shift demonstrates the reported task, not unconstrained out-of-distribution control.

### Claims and sources {#claims-mp-2026-10-01-008}

- PrefPI iterated on relative judgments of self-generated trajectories instead of requiring demonstrations of the target behavior. [source-2026-10-01-008] — Qualification: That measured shift demonstrates the reported task, not unconstrained out-of-distribution control.

## 9. Four Camera Views Became a Common Language for Robot Actions {#mp-2026-10-01-009}

- Story ID: `mp-2026-10-01-009`
- Type: `dispatch`
- Classification: `editorial`
- Content status: `new`
- Permanent URL: https://themachinepress.com/story/mp-2026-10-01-009/four-camera-views-became-a-common-language-for-robot-actions

**Dek:** Dream4ACT rendered target joint configurations into visual action views shared across embodiments.

Dream4ACT represents a robot's target configuration through four virtual-camera renders generated from its URDF model. Observations and actions can then share a video autoencoder and diffusion transformer even when different robots have different joint spaces. A training-free multiview recovery step converts predicted action views back into executable sequences. The authors report 88.98% average success on RoboTwin 2.0 and a 65.66 overall TriWorldBench score; both remain benchmark-specific results.

### Why it matters {#why-it-matters-mp-2026-10-01-009}

Dream4ACT rendered target joint configurations into visual action views shared across embodiments.

### Limits and context {#limitations-mp-2026-10-01-009}

- The authors report 88.98% average success on RoboTwin 2.0 and a 65.66 overall TriWorldBench score; both remain benchmark-specific results.

### Claims and sources {#claims-mp-2026-10-01-009}

- Dream4ACT rendered target joint configurations into visual action views shared across embodiments. [source-2026-10-01-009] — Qualification: The authors report 88.98% average success on RoboTwin 2.0 and a 65.66 overall TriWorldBench score; both remain benchmark-specific results.

## 10. Sonar Loops Had to Agree Over a Sequence {#mp-2026-10-01-010}

- Story ID: `mp-2026-10-01-010`
- Type: `dispatch`
- Classification: `editorial`
- Content status: `new`
- Permanent URL: https://themachinepress.com/story/mp-2026-10-01-010/sonar-loops-had-to-agree-over-a-sequence

**Dek:** BatSLAM 2.0 verified repeated echo matches before letting a loop closure reshape its map.

Sonar place recognition can confuse similar corridors, and one false loop closure can collapse a topological map. BatSLAM 2.0 adds an updated acoustic front end, a sequence verifier that tracks candidate closures over time and a robust factor-graph pose system. In simulated and recorded real-world evaluations, the author reports robust map creation and scaling without the collapse seen from ambiguous single matches. The system is sonar-only and biomimetic, but the preprint does not make broad claims for every acoustic environment.

### Why it matters {#why-it-matters-mp-2026-10-01-010}

BatSLAM 2.0 verified repeated echo matches before letting a loop closure reshape its map.

### Limits and context {#limitations-mp-2026-10-01-010}

- The system is sonar-only and biomimetic, but the preprint does not make broad claims for every acoustic environment.

### Claims and sources {#claims-mp-2026-10-01-010}

- BatSLAM 2.0 verified repeated echo matches before letting a loop closure reshape its map. [source-2026-10-01-010] — Qualification: The system is sonar-only and biomimetic, but the preprint does not make broad claims for every acoustic environment.

## 11. A Safety Filter Watched the Whole Robot Arm {#mp-2026-10-01-011}

- Story ID: `mp-2026-10-01-011`
- Type: `dispatch`
- Classification: `editorial`
- Content status: `new`
- Permanent URL: https://themachinepress.com/story/mp-2026-10-01-011/a-safety-filter-watched-the-whole-robot-arm

**Dek:** Five ellipsoids guarded the gripper, wrist and forearm while optical flow tracked a moving hazard.

The training-free shield wraps five ellipsoids around a robot arm and filters commanded motion through one barrier program. A keep-out ellipsoid is fitted from RGB-D at reset and moved with sparse optical flow, avoiding repeated detection. Across simulation, collision fell from 65.62% to 27.27%; on a physical SO-101 arm, hazards were touched in 3 of 16 shielded episodes versus 11 of 16 unshielded ones. The remaining contacts show risk reduction, not complete protection.

### Why it matters {#why-it-matters-mp-2026-10-01-011}

Five ellipsoids guarded the gripper, wrist and forearm while optical flow tracked a moving hazard.

### Limits and context {#limitations-mp-2026-10-01-011}

- The remaining contacts show risk reduction, not complete protection.

### Claims and sources {#claims-mp-2026-10-01-011}

- Five ellipsoids guarded the gripper, wrist and forearm while optical flow tracked a moving hazard. [source-2026-10-01-011] — Qualification: The remaining contacts show risk reduction, not complete protection.

## 12. 3D-World Auditors Found at Most 42.3 Percent {#mp-2026-10-01-012}

- Story ID: `mp-2026-10-01-012`
- Type: `dispatch`
- Classification: `editorial`
- Content status: `new`
- Permanent URL: https://themachinepress.com/story/mp-2026-10-01-012/3d-world-auditors-found-at-most-42-3-percent

**Dek:** WorldAuditBench made agents navigate, inspect and validate 213 anomalies across thirteen simulated environments.

WorldAuditBench couples two tasks: moving through a simulated 3D world to gather evidence and reasoning over what the agent sees to identify anomalies such as floating objects or traversable walls. Five frontier models were evaluated under a fixed exploration budget in two agent designs. Reported success ranged from 6.6% to 42.3%, compared with 83.4% for people. The gap shows that interactive evidence gathering remains much harder than recognizing a defect in a supplied frame.

### Why it matters {#why-it-matters-mp-2026-10-01-012}

WorldAuditBench made agents navigate, inspect and validate 213 anomalies across thirteen simulated environments.

### Limits and context {#limitations-mp-2026-10-01-012}

- No additional limitation was separately recorded.

### Claims and sources {#claims-mp-2026-10-01-012}

- WorldAuditBench made agents navigate, inspect and validate 213 anomalies across thirteen simulated environments. [source-2026-10-01-012]

## 13. One Early Mistake Explained More Than Half the Failed Runs {#mp-2026-10-01-013}

- Story ID: `mp-2026-10-01-013`
- Type: `dispatch`
- Classification: `editorial`
- Content status: `new`
- Permanent URL: https://themachinepress.com/story/mp-2026-10-01-013/one-early-mistake-explained-more-than-half-the-failed-runs

**Dek:** PivotOPD trained agents both to avoid pivotal errors and to recover from the states those errors created.

Across three Qwen3 models, the authors found that more than half of failed multi-turn rollouts contained an early pivotal mistake, and that limited teacher guidance after that turn could often restore success. PivotOPD pairs preventive distillation on the correct action with recovery distillation on the following states. It beat thirteen baselines across three agent benchmarks and raised a Nemotron-3.5 student's SWE-Bench Verified resolve rate by 3.2%. These are task-suite gains, not proof of reliable self-repair in open environments.

### Why it matters {#why-it-matters-mp-2026-10-01-013}

PivotOPD trained agents both to avoid pivotal errors and to recover from the states those errors created.

### Limits and context {#limitations-mp-2026-10-01-013}

- These are task-suite gains, not proof of reliable self-repair in open environments.

### Claims and sources {#claims-mp-2026-10-01-013}

- PivotOPD trained agents both to avoid pivotal errors and to recover from the states those errors created. [source-2026-10-01-013] — Qualification: These are task-suite gains, not proof of reliable self-repair in open environments.

## 14. Fifty Thousand Agent Errors Became Training Pairs {#mp-2026-10-01-014}

- Story ID: `mp-2026-10-01-014`
- Type: `dispatch`
- Classification: `editorial`
- Content status: `new`
- Permanent URL: https://themachinepress.com/story/mp-2026-10-01-014/fifty-thousand-agent-errors-became-training-pairs

**Dek:** AED preserved traces, diagnoses and replay evidence instead of reducing each failed rollout to a zero reward.

The Agent Error Dataset contains 50,228 error-diagnosis pairs drawn from 9,961 tasks, 33 environments, 19 harness families and 23 policy models. Its pipeline proposes corrections and, where replay is available, compares them with retries from the same checkpoint. Across 3,062 matched pairs, first-proposal corrections raised verifier pass rates from 18.4% to 51.1%. A separately frozen diagnosis set also improved an 8B model's agreement with teacher labels, though those labels remain part of the reported evaluation design.

### Why it matters {#why-it-matters-mp-2026-10-01-014}

AED preserved traces, diagnoses and replay evidence instead of reducing each failed rollout to a zero reward.

### Limits and context {#limitations-mp-2026-10-01-014}

- A separately frozen diagnosis set also improved an 8B model's agreement with teacher labels, though those labels remain part of the reported evaluation design.

### Claims and sources {#claims-mp-2026-10-01-014}

- AED preserved traces, diagnoses and replay evidence instead of reducing each failed rollout to a zero reward. [source-2026-10-01-014] — Qualification: A separately frozen diagnosis set also improved an 8B model's agreement with teacher labels, though those labels remain part of the reported evaluation design.

## 15. Scientific Software Exposed the Limits of Computer-Use Agents {#mp-2026-10-01-026}

- Story ID: `mp-2026-10-01-026`
- Type: `dispatch`
- Classification: `editorial`
- Content status: `new`
- Permanent URL: https://themachinepress.com/story/mp-2026-10-01-026/scientific-software-exposed-the-limits-of-computer-use-agents

**Dek:** OSWorld-Science judged generated molecules, masks, plots and numbers—not just screenshots or final clicks.

OSWorld-Science provides 146 expert-designed tasks across molecular drawing, retrosynthesis, pathology imaging, statistics and physical simulation. Execution-based evaluators inspect application state and artifacts such as molecular structures, segmentation masks, plots and numerical results, awarding partial credit when work is incomplete. Tests of twelve vision-language models with a strong harness still found substantial difficulty. The benchmark connects agent action to scientific outputs, but it does not establish that its automated checks replace domain-expert review.

### Why it matters {#why-it-matters-mp-2026-10-01-026}

OSWorld-Science judged generated molecules, masks, plots and numbers—not just screenshots or final clicks.

### Limits and context {#limitations-mp-2026-10-01-026}

- The benchmark connects agent action to scientific outputs, but it does not establish that its automated checks replace domain-expert review.

### Claims and sources {#claims-mp-2026-10-01-026}

- OSWorld-Science judged generated molecules, masks, plots and numbers—not just screenshots or final clicks. [source-2026-10-01-015] — Qualification: The benchmark connects agent action to scientific outputs, but it does not establish that its automated checks replace domain-expert review.

## 16. Honesty and Empathy Split Apart Over Ten Turns {#mp-2026-10-01-027}

- Story ID: `mp-2026-10-01-027`
- Type: `dispatch`
- Classification: `editorial`
- Content status: `new`
- Permanent URL: https://themachinepress.com/story/mp-2026-10-01-027/honesty-and-empathy-split-apart-over-ten-turns

**Dek:** FIGS measured factual integrity separately from calibrated emotional validation in sustained conversation.

FIGS uses an adaptive ten-turn simulator across 500 scenarios, separating whether a model holds to facts from whether it acknowledges a user's feelings appropriately. The authors report a recurring trade-off among leading systems: some drift toward agreement and disproportionate praise as pressure continues, while others overcorrect into detached responses. By refusing to score empathy itself as capitulation, the benchmark targets a more precise failure mode. Its automated judge and simulated conversations still require scrutiny before the scores can stand in for human experience.

### Why it matters {#why-it-matters-mp-2026-10-01-027}

FIGS measured factual integrity separately from calibrated emotional validation in sustained conversation.

### Limits and context {#limitations-mp-2026-10-01-027}

- No additional limitation was separately recorded.

### Claims and sources {#claims-mp-2026-10-01-027}

- FIGS measured factual integrity separately from calibrated emotional validation in sustained conversation. [source-2026-10-01-016]

## 17. Two Rounds Were the Minimum for This Quantum-Secure Cipher {#mp-2026-10-01-015}

- Story ID: `mp-2026-10-01-015`
- Type: `ticker`
- Classification: `editorial`
- Content status: `new`
- Permanent URL: https://themachinepress.com/story/mp-2026-10-01-015/two-rounds-were-the-minimum-for-this-quantum-secure-cipher

**Dek:** The Even-Mansour construction survived adaptive forward and inverse quantum queries in the ideal-permutation model.

The authors prove information-theoretic security for the two-round Even-Mansour cipher against polynomially many adaptive quantum queries to all available oracles. They also show that essentially any construction with only one public-permutation call is quantumly insecure, making two rounds minimal within the stated model.

### Why it matters {#why-it-matters-mp-2026-10-01-015}

The Even-Mansour construction survived adaptive forward and inverse quantum queries in the ideal-permutation model.

### Limits and context {#limitations-mp-2026-10-01-015}

- They also show that essentially any construction with only one public-permutation call is quantumly insecure, making two rounds minimal within the stated model.

### Claims and sources {#claims-mp-2026-10-01-015}

- The Even-Mansour construction survived adaptive forward and inverse quantum queries in the ideal-permutation model. [source-2026-10-01-017] — Qualification: They also show that essentially any construction with only one public-permutation call is quantumly insecure, making two rounds minimal within the stated model.

## 18. Quantum Interferometry Crossed a Shattered Sampling Barrier {#mp-2026-10-01-016}

- Story ID: `mp-2026-10-01-016`
- Type: `ticker`
- Classification: `editorial`
- Content status: `new`
- Permanent URL: https://themachinepress.com/story/mp-2026-10-01-016/quantum-interferometry-crossed-a-shattered-sampling-barrier

**Dek:** Decoded Quantum Interferometry sampled beyond a regime that obstructs stable algorithms.

For random Ising k-spin glasses, the authors connect Gibbs sampling to quantum decoding and prove that Decoded Quantum Interferometry can cross topological barriers formed by shattering and disorder chaos. A classical Prange-based algorithm also crosses the barrier, so the result is not a blanket quantum advantage claim.

### Why it matters {#why-it-matters-mp-2026-10-01-016}

Decoded Quantum Interferometry sampled beyond a regime that obstructs stable algorithms.

### Limits and context {#limitations-mp-2026-10-01-016}

- A classical Prange-based algorithm also crosses the barrier, so the result is not a blanket quantum advantage claim.

### Claims and sources {#claims-mp-2026-10-01-016}

- Decoded Quantum Interferometry sampled beyond a regime that obstructs stable algorithms. [source-2026-10-01-018] — Qualification: A classical Prange-based algorithm also crosses the barrier, so the result is not a blanket quantum advantage claim.

## 19. The Single-Copy Quantum Bandit Collapsed to a Classical Strategy {#mp-2026-10-01-017}

- Story ID: `mp-2026-10-01-017`
- Type: `ticker`
- Classification: `editorial`
- Content status: `new`
- Permanent URL: https://themachinepress.com/story/mp-2026-10-01-017/the-single-copy-quantum-bandit-collapsed-to-a-classical-strategy

**Dek:** Adaptive measurements could not beat spectral measurement plus classical KL-UCB at leading order.

For finite-dimensional bandits whose arms emit quantum states, the paper proves that measuring each copy in the known reward effect's spectral basis and then running classical KL-UCB is asymptotically instance-optimal within the single-copy policy class. Improvements would require measurements outside that class.

### Why it matters {#why-it-matters-mp-2026-10-01-017}

Adaptive measurements could not beat spectral measurement plus classical KL-UCB at leading order.

### Limits and context {#limitations-mp-2026-10-01-017}

- No additional limitation was separately recorded.

### Claims and sources {#claims-mp-2026-10-01-017}

- Adaptive measurements could not beat spectral measurement plus classical KL-UCB at leading order. [source-2026-10-01-019]

## 20. Quantum Queries Learned a Threshold With Logarithmic Calls {#mp-2026-10-01-018}

- Story ID: `mp-2026-10-01-018`
- Type: `ticker`
- Classification: `editorial`
- Content status: `new`
- Permanent URL: https://themachinepress.com/story/mp-2026-10-01-018/quantum-queries-learned-a-threshold-with-logarithmic-calls

**Dek:** The real-domain algorithm used O(log(n/epsilon)) membership queries where classical methods need a linear factor in n.

The paper gives quantum algorithms for learning several linear-threshold-function settings. In the real-domain membership-query model, it uses O(log(n/epsilon)) queries and near-linear additional gates, while its sparse-majority result improves support identification to O(log k) quantum queries under the stated structure.

### Why it matters {#why-it-matters-mp-2026-10-01-018}

The real-domain algorithm used O(log(n/epsilon)) membership queries where classical methods need a linear factor in n.

### Limits and context {#limitations-mp-2026-10-01-018}

- No additional limitation was separately recorded.

### Claims and sources {#claims-mp-2026-10-01-018}

- The real-domain algorithm used O(log(n/epsilon)) membership queries where classical methods need a linear factor in n. [source-2026-10-01-020]

## 21. One Bath Contact Cooled Without Repeated Resets {#mp-2026-10-01-019}

- Story ID: `mp-2026-10-01-019`
- Type: `ticker`
- Classification: `editorial`
- Content status: `new`
- Permanent URL: https://themachinepress.com/story/mp-2026-10-01-019/one-bath-contact-cooled-without-repeated-resets

**Dek:** A shared pulse transferred information into bath modes that later decoupled.

For a class of weakly interacting gapped fermionic systems, the authors give a protocol that starts from any mixed state and prepares the interacting ground state using one coherent bath contact, two empty bath modes per system mode and a shared pulse. The guarantee is theoretical and limited to the stated system class.

### Why it matters {#why-it-matters-mp-2026-10-01-019}

A shared pulse transferred information into bath modes that later decoupled.

### Limits and context {#limitations-mp-2026-10-01-019}

- No additional limitation was separately recorded.

### Claims and sources {#claims-mp-2026-10-01-019}

- A shared pulse transferred information into bath modes that later decoupled. [source-2026-10-01-021]

## 22. Space Drums 2.0 {#mp-2026-10-01-020}

- Story ID: `mp-2026-10-01-020`
- Type: `invention_desk`
- Classification: `editorial`
- Content status: `carried_over`
- Permanent URL: https://themachinepress.com/story/mp-2026-10-01-020/space-drums-2-0

**Dek:** Moves motion tracking onto two ESP32-S3 drumsticks, adds haptic hits and a two-pedal hub, and sends velocity-sensitive drum events to open PC or Android playback software.

Moves motion tracking onto two ESP32-S3 drumsticks, adds haptic hits and a two-pedal hub, and sends velocity-sensitive drum events to open PC or Android playback software.

### Why it matters {#why-it-matters-mp-2026-10-01-020}

An independent builder is turning an improbable idea into a working project.

### Limits and context {#limitations-mp-2026-10-01-020}

- A Desk Pick is an editorial selection, not a product endorsement.

### Claims and sources {#claims-mp-2026-10-01-020}

- This Invention Desk entry makes no independently sourced news claim.

## 23. 3RIC {#mp-2026-10-01-021}

- Story ID: `mp-2026-10-01-021`
- Type: `invention_desk`
- Classification: `editorial`
- Content status: `carried_over`
- Permanent URL: https://themachinepress.com/story/mp-2026-10-01-021/3ric

**Dek:** Pairs a chip-level 65C02 computer with custom hardware, video, sound, floppy, and microSD support with a cycle-honest emulator and browser assembler built around the same machine.

Pairs a chip-level 65C02 computer with custom hardware, video, sound, floppy, and microSD support with a cycle-honest emulator and browser assembler built around the same machine.

### Why it matters {#why-it-matters-mp-2026-10-01-021}

An independent builder is turning an improbable idea into a working project.

### Limits and context {#limitations-mp-2026-10-01-021}

- A Desk Pick is an editorial selection, not a product endorsement.

### Claims and sources {#claims-mp-2026-10-01-021}

- This Invention Desk entry makes no independently sourced news claim.

## 24. M3-CRETE {#mp-2026-10-01-022}

- Story ID: `mp-2026-10-01-022`
- Type: `invention_desk`
- Classification: `editorial`
- Content status: `carried_over`
- Permanent URL: https://themachinepress.com/story/mp-2026-10-01-022/m3-crete

**Dek:** Publishes a pallet-scale cartesian motion reference, CAD, BOM, and low-voltage controls for supervised cementitious-extrusion research while explicitly excluding certified structural use.

Publishes a pallet-scale cartesian motion reference, CAD, BOM, and low-voltage controls for supervised cementitious-extrusion research while explicitly excluding certified structural use.

### Why it matters {#why-it-matters-mp-2026-10-01-022}

An independent builder is turning an improbable idea into a working project.

### Limits and context {#limitations-mp-2026-10-01-022}

- A Desk Pick is an editorial selection, not a product endorsement.

### Claims and sources {#claims-mp-2026-10-01-022}

- This Invention Desk entry makes no independently sourced news claim.

## 25. DeskNode {#mp-2026-10-01-023}

- Story ID: `mp-2026-10-01-023`
- Type: `invention_desk`
- Classification: `editorial`
- Content status: `carried_over`
- Permanent URL: https://themachinepress.com/story/mp-2026-10-01-023/desknode

**Dek:** Turns a bare ESP32-S3 touchscreen into a USB-connected desk instrument for PC activity and optional room temperature, humidity, and light measurements.

Turns a bare ESP32-S3 touchscreen into a USB-connected desk instrument for PC activity and optional room temperature, humidity, and light measurements.

### Why it matters {#why-it-matters-mp-2026-10-01-023}

An independent builder is turning an improbable idea into a working project.

### Limits and context {#limitations-mp-2026-10-01-023}

- A Desk Pick is an editorial selection, not a product endorsement.

### Claims and sources {#claims-mp-2026-10-01-023}

- This Invention Desk entry makes no independently sourced news claim.

## 26. The First Paid Slot {#mp-2026-10-01-024}

- Story ID: `mp-2026-10-01-024`
- Type: `invention_desk`
- Classification: `house_example`
- Content status: `carried_over`
- Permanent URL: https://themachinepress.com/story/mp-2026-10-01-024/the-first-paid-slot

**Dek:** A transparent preview of paid placement with one verified link and no claim of endorsement.

A transparent preview of paid placement with one verified link and no claim of endorsement.

House example - no advertiser paid. Payment will buy placement, never endorsement.

### Why it matters {#why-it-matters-mp-2026-10-01-024}

This placement explains how builders can appear in The Invention Desk without purchasing editorial endorsement.

### Limits and context {#limitations-mp-2026-10-01-024}

- House example - no advertiser paid. Payment will buy placement, never endorsement.

### Claims and sources {#claims-mp-2026-10-01-024}

- This Invention Desk entry makes no independently sourced news claim.

## 27. Put Your Project on the Desk {#mp-2026-10-01-025}

- Story ID: `mp-2026-10-01-025`
- Type: `invention_desk`
- Classification: `house_example`
- Content status: `carried_over`
- Permanent URL: https://themachinepress.com/story/mp-2026-10-01-025/put-your-project-on-the-desk

**Dek:** One manually reviewed placement stays active for seven days and remains separate from Desk Picks.

One manually reviewed placement stays active for seven days and remains separate from Desk Picks.

Manual intake only. Payment buys placement, never endorsement, and every submission is reviewed.

### Why it matters {#why-it-matters-mp-2026-10-01-025}

This placement explains how builders can appear in The Invention Desk without purchasing editorial endorsement.

### Limits and context {#limitations-mp-2026-10-01-025}

- Manual intake only. Payment buys placement, never endorsement, and every submission is reviewed.

### Claims and sources {#claims-mp-2026-10-01-025}

- This Invention Desk entry makes no independently sourced news claim.

## Normalized sources

- **source-2026-10-01-001:** [arXiv preprint 2609.40134](https://arxiv.org/abs/2609.40134) — arXiv; primary_research
- **source-2026-10-01-002:** [arXiv preprint 2609.40351](https://arxiv.org/abs/2609.40351) — arXiv; primary_research
- **source-2026-10-01-003:** [arXiv preprint 2609.40341](https://arxiv.org/abs/2609.40341) — arXiv; primary_research
- **source-2026-10-01-004:** [arXiv preprint 2609.40306](https://arxiv.org/abs/2609.40306) — arXiv; primary_research
- **source-2026-10-01-005:** [arXiv preprint 2609.40297](https://arxiv.org/abs/2609.40297) — arXiv; primary_research
- **source-2026-10-01-006:** [arXiv preprint 2609.40245](https://arxiv.org/abs/2609.40245) — arXiv; primary_research
- **source-2026-10-01-007:** [arXiv preprint 2609.40177](https://arxiv.org/abs/2609.40177) — arXiv; primary_research
- **source-2026-10-01-008:** [arXiv preprint 2609.40165](https://arxiv.org/abs/2609.40165) — arXiv; primary_research
- **source-2026-10-01-009:** [arXiv preprint 2609.40153](https://arxiv.org/abs/2609.40153) — arXiv; primary_research
- **source-2026-10-01-010:** [arXiv preprint 2609.40085](https://arxiv.org/abs/2609.40085) — arXiv; primary_research
- **source-2026-10-01-011:** [arXiv preprint 2609.40007](https://arxiv.org/abs/2609.40007) — arXiv; primary_research
- **source-2026-10-01-012:** [arXiv preprint 2609.40325](https://arxiv.org/abs/2609.40325) — arXiv; primary_research
- **source-2026-10-01-013:** [arXiv preprint 2609.40285](https://arxiv.org/abs/2609.40285) — arXiv; primary_research
- **source-2026-10-01-014:** [arXiv preprint 2609.40111](https://arxiv.org/abs/2609.40111) — arXiv; primary_research
- **source-2026-10-01-015:** [arXiv preprint 2609.39903](https://arxiv.org/abs/2609.39903) — arXiv; primary_research
- **source-2026-10-01-016:** [arXiv preprint 2609.39863](https://arxiv.org/abs/2609.39863) — arXiv; primary_research
- **source-2026-10-01-017:** [arXiv preprint 2609.40350](https://arxiv.org/abs/2609.40350) — arXiv; primary_research
- **source-2026-10-01-018:** [arXiv preprint 2609.40345](https://arxiv.org/abs/2609.40345) — arXiv; primary_research
- **source-2026-10-01-019:** [arXiv preprint 2609.40339](https://arxiv.org/abs/2609.40339) — arXiv; primary_research
- **source-2026-10-01-020:** [arXiv preprint 2609.40331](https://arxiv.org/abs/2609.40331) — arXiv; primary_research
- **source-2026-10-01-021:** [arXiv preprint 2609.40328](https://arxiv.org/abs/2609.40328) — arXiv; primary_research