Carlos's Debrief

June 11, 2026 04:00
46ArXiv Papers
48Web Findings
94Total Sources
Topics searched: Artificial Intelligence Machine Learning Large Language Models Security & Cybersecurity Cryptography Zero Knowledge Quantum Computing Crypto & Blockchain AI Agents & Reasoning AI Safety & Alignment
June 11, 2026 04:00 — Curated by Hermes Research Scout

⚡ Quick Summary

📰 News Headlines

📄 ArXiv Papers

🧠 Reroute / Don / Remove 4

Reroute, Don't Remove: Recoverable Visual Token Routing for Vision-Language Models [1]
Cheng-Yu Yang, Shao-Yuan Lo, Yu-Lun Liu
Vision-language models (VLMs) project images into hundreds to thousands of visual tokens, making decoder inference expensive in both attention computation and KV-cache memory. Existing visual-token…
cs.CVcs.AI
DIRECT: When and Where Should You Allocate Test-Time Compute in Embodied Planners? [3]
Jadelynn Dao, Milan Ganai, Yasmina Abukhadra, Ajay Sridhar, Mozhgan Nasr Azadani, Katie Luo, Clark Barrett, Jiajun Wu, Chelsea Finn, Marco Pavone
Vision-Language Models (VLMs) are increasingly deployed as high-level planners for embodied agents, with an emerging strategy of scaling test-time compute to improve capability. However, we observe…
cs.ROcs.AIcs.CV
World Pilot: Steering Vision-Language-Action Models with World-Action Priors [20]
Zefu Lin, Rongxu Cui, Junjia Xu, Xiaojuan Jin, Wenling Li, Lue Fan, Zhaoxiang Zhang
Vision-Language-Action (VLA) models inherit semantic grounding from large-scale pretraining and perform competently across in-distribution manipulation tasks. This grounding, however, is built on…
cs.RO
VLGA: Vision-Language-Geometry-Action Models for Autonomous Driving [44]
Jin Yao, Dhruva Dixith Kurra, Tom Lampo, Zezhou Cheng, Danhua Guo, Burhan Yaman
Vision-language-action (VLA) models can describe scenes and reason about them in language, yet still struggle to ground their actions in the dense 3D world around them. Existing approaches either…
cs.CVcs.RO

🧠 Agentic / Optimization / Reinforcement 4

Reinforcement Learning Disrupts Gradient-Based Adversarial Optimization [28]
Xinhai Zou, Chang Zhao, Alireza Aghabagherloo, Dave Singelée, Robin Degraeve, Bart Preneel
Gradient-based adversarial attacks remain a dominant threat to deep neural networks (DNNs), as they exploit gradient information to efficiently optimize adversarial perturbations. To address this, we…
cs.LGcs.AIcs.CR
Breaking Entropy Bounds: Accelerating RL Training via MTP with Rejection Sampling [13]
Yucheng Li, Huiqiang Jiang, Yang Xu, Jianxin Yang, Yi Zhang, Yizhong Cao, Yuhao Shen, Fan Zhou, Rui Men, Jianwei Zhang, An Yang, Bowen Yu, Bo Zheng, Fei Huang, Junyang Lin, Dayiheng Liu, Jingren Zhou
Reinforcement learning (RL) has become a key component in modern large language models, yet the rollout stage remains the key bottleneck in RL training pipelines. Although Multi-Token Prediction…
cs.LGcs.CL
APPO: Agentic Procedural Policy Optimization [8]
Xucong Wang, Ziyu Ma, Yong Wang, Yuxiang Ji, Shidong Yang, Guanhua Chen, Pengkun Wang, Xiangxiang Chu
Recent advances in agentic Reinforcement Learning (RL) have substantially improved the multi-turn tool-use capabilities of large language model agents. However, most existing methods assign credit…
cs.LGcs.AI
UniIntervene: Agentic Intervention for Efficient Real-World Reinforcement Learning [12]
Haoyuan Deng, Yitong Gao, Yudong Lin, Haichao Liu, Zhenyu Wu, Ziwei Wang
Human-in-the-loop reinforcement learning (HiL-RL) has emerged as an effective paradigm for real-world robotic manipulation, enabling online policy improvement with human guidance. However, current…
cs.ROcs.LG

🧠 Context-driven / Incremental / Compression 2

Doc-to-Atom: Learning to Compile and Compose Memory Atoms [35]
Xingjian Diao, Wenbo Li, Yashas Malur Saidutta, Avinash Amballa, Lazar Valkov, Srinivas Chappidi
Long input sequences are central to document understanding and multi-step reasoning in Large Language Models, yet the quadratic cost of attention makes inference both memory-intensive and slow.…
cs.CLcs.IR
Context-Driven Incremental Compression for Multi-Turn Dialogue Generation [11]
Yeongseo Jung, Jaehyeok Kim, Eunseo Jung, Jiachuan Wang, Yongqi Zhang, Ka Chun Cheung, Simon See, Lei Chen
Modern conversational agents condition on an ever-growing dialogue history at each turn, incurring redundant attention and encoding costs that grow with conversation length. Naive truncation or…
cs.CLcs.LG

🧠 Bounding / Effect / Hod 2

Bounding the Effect of HOD Assumptions on Small-Scale Clustering Constraints [19]
Nick Magnelli, Zachery Brown, Lado Samushia
Small-scale galaxy clustering is expected to contain substantial cosmological information, but the extent to which this information constrains halo-based cosmologies independent of an assumed…
astro-ph.CO
KiDS-Legacy: Joint analysis of second- and third-order cosmic shear [45]
L. Linke, L. Porth, P. Burger, J. Harnois-Déraps, S. Heydenreich, P. Schneider, M. Asgari, M. Bilicki, C. Georgiou, C. Heymans, H. Hildebrandt, H. Hoekstra, P. Jalan, B. Joachimi, S. Joudaki, K. Kuijken, S. Li, L. Moscardini, M. Radovich, R. Reischke, B. Stölzner, A. H. Wright, Z. Yan, Y. -H. Zhang
Weak lensing by large-scale structure is a powerful cosmological probe. While most analyses rely on second-order correlations, these are primarily sensitive to the parameter combination $S_8 = σ_8…
astro-ph.CO

🧠 Llm / Ocelot / Inference-leakage 2

OCELOT: Inference-Leakage Budgets for Privacy-Preserving LLM Agents [24]
Jin Xie, Songze Li
Large language model (LLM) agents increasingly act on a user's behalf -- reading personal files, calling tools, transacting with external services -- possibly leaking personally identifiable…
cs.CR
Selection Integrity for LLM Graph Memory: An Accumulability Criterion for Information-Flow-Blind Retrieval [26]
Zeming Fei, Hongming Fei, Xiaoyang Wang, Yang yang, Prosanta Gope, Biplab Sikdar, Ying Zhang
Agent memory is moving to graphs, and the provenance defenses now being built for it all check one thing: the provenance of the records an agent retrieves. We show that this entire class of defense…
cs.CR

🧠 Homomorphic / Jaguar / Fast 2

WHET: Welding Homomorphic Encryption to Accelerator Architectures [34]
Jongmin Kim, Hyesung Ji, Wonseok Choi, Hyunah Yu, Jung Ho Ahn
Fully homomorphic encryption (FHE) enables computations on encrypted data without decryption, offering strong data privacy at the expense of substantial computational and memory overheads. Prior…
cs.CR
Jaguar: Fast Private CNN Inference with Power-of-Two Homomorphic Arithmetic [33]
Yewon Jeong, Nayoung Jung, Hyeri Roh, Woo-Seok Choi
Hybrid HE/2PC private CNN inference remains bottlenecked by prime-modulus homomorphic arithmetic in convolution and by a precision flow that runs ReLU at doubled bitwidth before invoking a separate…
cs.CR

🧠 Factr / External / Force 1

FACTR 2: Learning External Force Sensing for Commodity Robot Arms Improves Policy Learning [2]
Steven Oh, Jason Jingzhou Liu, Tony Tao, Philip Han, Kenneth Shaw, Satoshi Funabashi, Ruslan Salakhutdinov, Deepak Pathak
Contact-rich manipulation requires force sensitivity, but many robot arms lack dedicated force sensors due to their high cost. We present Neural External Torque Estimation (NEXT), a data-driven…
cs.ROcs.AIcs.LG

🧠 Redesign / Mixture-of-experts / Routers 1

Redesign Mixture-of-Experts Routers with Manifold Power Iteration [4]
Songhao Wu, Ang Lv, Ruobing Xie, Yankai Lin
Router is the cornerstone component to the Mixture-of-Experts models. Serving as expert proxies, the rows of the router matrix compute their similarity to the MoE inputs to determine which subset of…
cs.LGcs.AIcs.CL

🧠 System / Report / Ccl25-eval 1

System Report for CCL25-Eval Task 5: New Dataset and LoRA-Fine-Tuned Qwen2.5 [5]
Haotao Xie
Recently, large language models (LLMs) have achieved promising progress in the fields of classical Chinese translation and the generation of classical poetry. However, domain-specific research on…
cs.CLcs.AI

🧠 Tahoe / Text-to-sql / Automated 1

TAHOE: Text-to-SQL with Automated Hint Optimization from Experience [6]
Zhiyi Chen, Jie Song, Peng Li
Large Language Models (LLMs) have democratized database access through Text-to-SQL, but moving from prototypes to production remains difficult. Real deployments must handle strict SQL dialects,…
cs.DBcs.AI

🧠 Atlas / Active / Theory 1

ATLAS: Active Theory Learning for Automated Science [7]
Noémi Éltető, Nathaniel D. Daw, Kimberly L. Stachenfeld, Kevin J. Miller
Advancing scientific understanding through mechanistic modeling requires posing the right experimental questions to yield maximally informative data. To automate this pursuit within cognitive…
cs.LGcs.AI

🧠 Spea2 / Improved / Density 1

SPEA2$^+$: Improved Density Estimation in SPEA2 with Provable Runtime Guarantees [9]
Duc-Cuong Dang, Andre Opris, Dirk Sudholt
The Strength Pareto Evolutionary Algorithm 2 (SPEA2) is a popular and prominent evolutionary algorithm for solving multi-objective optimisation problems. Despite its popularity, theoretical analyses…
cs.NEcs.AI

🧠 Illumination-robust / Camera-based / Heart-rate 1

Illumination-Robust Camera-Based Heart-Rate Estimation for Physiological Sensing in Robots [10]
Zhi Wei Xu, Torbjörn E. M. Nordling
Physiological awareness is important for service, social, and assistive robots that interact with humans in everyday environments. Remote photoplethysmography (rPPG) enables non-contact heart-rate…
cs.CVcs.AI

🧠 Subquadratic / Architectures / Applications 1

On Subquadratic Architectures: From Applications to Principles [14]
Anamaria-Roberta Hartl, Levente Zólyomi, David Stap, Pieter-Jan Hoedt, Niklas Schmidinger, Lukas Hauzenberger, Sebastian Böck, Günter Klambauer, Sepp Hochreiter
Transformers dominate modern sequence modeling, but their quadratic attention incurs substantial computational cost. Subquadratic architectures offer a scalable alternative. However, it remains…
cs.LG

🧠 Latent / World / Recovery 1

Latent World Recovery for Multimodal Learning with Missing Modalities [15]
Hui Wang, Tianyu Ren, Joseph Butler, Christopher Baker, Karen Rafferty, Simon McDade
We study multimodal learning under missing modalities, with particular motivation from bioscience applications in which heterogeneous modalities are often only partially available when decisions need…
cs.LGcs.AI

🧠 Anatomy / Post-training / Interpretability 1

Anatomy of Post-Training: Using Interpretability to Characterize Data and Shape the Learning Signal [16]
Leon Bergen, Usha Bhalla, Sidharth Baskaran, Max Loeffler, Raphael Sarfati, Dhruvil Gala, Ryan Panwar, Santiago Aranguri, Thomas Fel, Atticus Geiger, Matthew Kowal, Siddharth Boppana, Daniel Balsam, Owen Lewis, Jack Merullo, Thomas McGrath, Ekdeep Singh Lubana
Language-model post-training is the main stage at which model behavior is shaped, yet it still largely involves optimization of scalar rewards that summarize diverse desiderata. This abstraction…
cs.LG

🧠 Freeze-in / When / Direct 1

When direct detection constrains reheating temperature: freeze-in with stronger couplings and inflaton-seeded freeze-in [17]
Xavier Bertou, Olivier Deligny, Mathieu Gross, Yann Mambrini, Issam-Eddine Mellouki
Recent results from the DAMIC-M and PandaX collaborations have excluded the standard freeze-in production of dark matter for masses in the range $3~\mathrm{MeV} \lesssim m_χ\lesssim 1~\mathrm{GeV}$…
hep-phastro-ph.CO

🧠 How / Seemingly / Inconsequential 1

How Seemingly Inconsequential Design Choices Dictate Performance of LLMs in Pathology [18]
Kian R. Weihrauch, Thomas A. Buckley, William Lotter, Arjun K. Manrai
General-purpose large language models (LLMs) are routinely used as baselines when evaluating specialized pathology models on whole-slide images (WSIs). Because WSIs exceed contemporary model context…
cs.CV

🧠 Carbon / Isotope / Ratio 1

The carbon isotope ratio of β Pic b with high-resolution spectroscopy [21]
D. González Picos, I. A. G. Snellen, R. Landman, S. de Regt, N. Grasser, J. L. Birkby, T. Stolker, I. Koutalios, M. A. Kenworthy
Isotopic ratios trace the formation and evolution of planets and link their atmospheres to the chemistry of their natal protoplanetary discs. We measure $^{12}\mathrm{C}/^{13}\mathrm{C} =…
astro-ph.EP

🧠 Marcim-wg / Cyber / Wargame 1

MARCIM-WG: A cyber wargame proposal based on math modeling applied in a naval scenario [22]
Diego Cabuya-Padilla, Daniel Díaz-López, Carlos Castaneda-Marroquín
As maritime operations increasingly depend on interconnected digital ecosystems, cyber incidents can propagate across maritime networks and degrade critical services. Strengthening strategic Cyber…
cs.CR

🧠 Ecysap / Eye / Cyber 1

ECYSAP EYE: From Cyber Situational Awareness to Mission-Centric Decision Support for Enhanced Cyberspace Operations [23]
Pantaleone Nespoli, Daniel Díaz-López, Sergio Lopez Bernal, Francisco Oliva Bermejo, Pedro González Megías, Jorge Maestre Vidal, Víctor Sobrino García, Gregorio Martínez Pérez
Operational organizations increasingly require Cyber Situational Awareness (CySA) capabilities that go beyond isolated technical alerts, providing mission-relevant artefacts that can be embedded into…
cs.CR

🧠 Five-plane / Reference / Architecture 1

A Five-Plane Reference Architecture for Runtime Governance of Production AI Agents [25]
Krti Tallam
Enterprise security was built to govern data boundaries: the protected surface was data at rest and in transit, and the controls -- access control, data-loss prevention, perimeter inspection --…
cs.AIcs.CCcs.CR

🧠 Partitioned / Tags / Shared 1

Partitioned Tags, Shared Data: Reconciling Strict Cache Isolation with Write-Shared Coherence [27]
Kartik Ramkrishnan, Stephen McCamant, Antonia Zhai, Pen Chung Yew
Cache partitioning is among the strongest structural defenses against eviction-based cache side channels, yet a decade-old design issue has blocked its widespread deployment in secure shared-OS…
cs.CRcs.AR

🧠 Bridging / Smart / City 1

Bridging the Smart City Cybersecurity Data Gap Through AI-Driven Synthetic Dataset Generation [29]
Stephanie Polczynski, John D. Hastings, Varghese Vaidyan, Kyle Korman
Smart cities rely on interconnected cyber-physical systems that integrate sensors, IoT devices, cloud platforms, and AI-driven services and decision-making. While these systems enhance city services,…
cs.CR

🧠 On-chip / Quantum / Randomness 1

On-Chip Quantum Randomness Amplification [30]
Lang Li, Yutian Wu, Giulio Chiribella, Ravishankar Ramanathan
Randomness amplification, the task of extracting uniform private bits from biased seeds that may be partly known by a malicious third party, is of central importance in cryptography. The highest…
quant-phphysics.optics

🧠 Relations / 24-th / Homotopy 1

Relations in the 24-th homotopy groups of spheres [31]
Toshiyuki Miyauchi, Juno Mukai
The main purpose of this note is to give a proof of the fact that the Toda brackets \ $\langle\barν,σ,\barν\rangle$ and $\langleν,η, \barσ\rangle$ are not trivial. This is an affirmative answer of…
math.AT

🧠 Neuro-relational / Programs / Unifying 1

Neuro-Relational Programs: Unifying Queries and Neural Computation over Structured Data [32]
Arie Soeteman, Balder ten Cate, Maurice Funk, Benny Kimelfeld, Carsten Lutz, Moritz Schönherr
The conventional approach to deep learning over relational databases applies neural models, such as Graph Neural Networks (GNNs), to a graph representation of the database. Recent approaches instead…
cs.DBcs.CCcs.LG

🧠 Note / Conjecture 1

A note on a conjecture of Ng [36]
Andrew Pearce-Crump
In this note we give a lower bound for the second moment of a ratio of zeta functions summed over the non-trivial zeros of the Riemann zeta function that is half the size of the conjectured value.…
math.NT

🧠 Depthmaster / Unified / Monocular 1

DepthMaster: Unified Monocular Depth Estimation for Perspective and Panoramic Images [37]
Pengfei Wang, Shihao Wang, Liyi Chen, Zhiyuan Ma, Guowen Zhang, Lei Zhang
While monocular depth estimation has achieved significant progress, achieving generalized metric depth estimation for both narrow field-of-view (FoV) perspectives and $360^\circ$ panoramas remains an…
cs.CV

🧠 Pfaffian / Quantum / Hall 1

A Pfaffian quantum Hall state of ultracold bosons [38]
Joyce Kwan, Perrin Segura, Yanfei Li, Tizian Blatz, Annie Zhi, Brice Bakkali-Hassani, Annabelle Bohrdt, Martin Greiter, Fabian Grusdt, Markus Greiner
Fractional quantum Hall states are a cornerstone of topological physics, hosting fractionally charged quasiparticles with exotic statistics that promise to enable topologically protected quantum…
cond-mat.quant-gascond-mat.str-elphysics.atom-ph

🧠 Collective / Neutrino / Oscillations 1

Collective neutrino oscillations: Many-body non-forward effects and non-classicality [39]
Julien Froustey, Ermal Rrapaj, Yuhao Liu, Gushu Li, Costin Iancu, Vincenzo Cirigliano
Neutrino evolution in dense astrophysical environments is typically described either within a quantum kinetic framework, which neglects the build-up of multi-body correlations, or through simplified…
hep-phastro-ph.HEquant-ph

🧠 Quantifying / Distribution / Biexciton 1

Quantifying the Distribution of Biexciton Emission Efficiencies in Colloidal Quantum Shells [40]
Tjom Arens, Dulanjan Harankahage, Divesh Nazar, Mikhail Zamkov, Freddy T. Rabouw
The efficiency of multi-photon emission is an important characteristic of quantum light sources. Bright multi-photon emission is desirable for high-power lighting and lasers, while its complete…
physics.opticscond-mat.mtrl-sci

🧠 Scaling-optimal / Purification / Noisy 1

Scaling-optimal purification of noisy qubit unitary channels [41]
Ryotaro Niwa, Satoshi Yoshida, Koki Ono, Takeru Utsumi, Zhaoyi Li, Yuxiang Yang, Ryuji Takagi, Mio Murao
We consider the problem of purifying noisy qubit unitary channels. Given the ability to apply an unknown qubit unitary channel followed by depolarizing noise, we aim to construct a superchannel that…
quant-ph

🧠 Markov / Property / Path 1

Markov property and path regularity for the solutions to SPDEs driven by cylindrical-martingale valued measures [42]
Santiago Cambronero, David Campos, C. A. Fonseca-Mora, Darío Mena
In this paper we prove the Markov property for the solution to stochastic partial differential equations driven by a cylindrical orthogonal martingale-valued measure. We assume our coefficients are…
math.PR

🧠 Fermions / Are / Fundamentally 1

Fermions are fundamentally more nonlocal than Bosons [43]
Fatemeh Moradi Kalarde, Sadra Boreiri, Xiangling Xu, Lucas Tendick, Salman Beigi, Paolo Perinotti, Tommaso Guaita, Marc-Olivier Renou
Bell's theorem shows that entangled quantum particles can exhibit correlations that classical particles cannot reproduce without an additional nonlocal resource, such as communication. In this sense,…
quant-ph

🧠 Simplified / Stabilizer / Zx-calculus 1

The Simplified Stabilizer ZX-Calculus is Minimal [46]
Harry K. Stoltz
The stabilizer fragment of the ZX calculus is amongst the most important fragments of the theory. The closely related Clifford+T fragment is approximately universal (arXiv:1705.11151). Additionally,…
quant-ph

🌐 Web Findings

🤗 HuggingFace Papers 21

Verifiable Environments Are LEGO Bricks: Recursive Composition for Reasoning Generalization [61]
Reinforcement Learning (RL) with verifiable environments has emerged as a powerful approach for enhancing the reasoning capabilities of Large Language Models (LLMs). While prior…
▲ 5HuggingFace Papers
Toward Generalist Autonomous Research via Hypothesis-Tree Refinement [59]
Scientific progress depends on a repeated loop of exploration, experimentation, and abstraction. Researchers test candidate directions, interpret the evidence, and carry the…
▲ 52HuggingFace Papers
ComBench: A Benchmark for Rigorous Proof Reasoning and Constructive Realization in Olympiad-Level Combinatorics [58]
Combinatorics is central to Olympiad-level mathematical problem solving, requiring deep discrete reasoning, creative constructions, and rigorous structural insight. Recent…
▲ 18HuggingFace Papers
ICA Lens: Interpreting Language Models Without Training Another Dictionary [70]
Finding interpretable directions in language-model representations is critical for understanding and controlling model behavior. Sparse autoencoders (SAEs) have become the…
▲ 14HuggingFace Papers
Distilling LLM Feedback for Lean Theorem Proving [54]
Post-training for reasoning models typically combines supervised fine-tuning with reinforcement learning from verifiable rewards, most commonly with GRPO. However, this algorithm…
HuggingFace Papers
DeNovoSWE: Scaling Long-Horizon Environments for Generating Entire Repositories from Scratch [69]
As the capabilities of LLM-based code agents continue to advance, their expected role is expanding beyond localized bug fixing in existing codebases toward architecting and…
▲ 24HuggingFace Papers
Agentic Environment Engineering for Large Language Models: A Survey of Environment Modeling, Synthesis, Evaluation, and Application [63]
Environments serve as interactive systems for large language model (LLM) based agents across diverse scenarios and play a crucial role in driving the continual evolution of model…
▲ 53HuggingFace Papers
EvoTrainer: Co-Evolving LLM Policies and Training Harnesses for Autonomous Agentic Reinforcement Learning [55]
Autonomous LLM training is often framed as recipe search, which leaves the training harness largely static. This limitation sharpens in agentic RL, where shifting bottlenecks and…
▲ 7HuggingFace Papers
On the Limits of LLM-as-Judge for Scientific Novelty Assessment [71]
LLMs are increasingly used to generate and judge scientific ideas. This makes novelty evaluation a central problem. Full idea evaluation is difficult because it often requires…
▲ 1HuggingFace Papers
World Model Self-Distillation: Training World Models to Solve General Tasks [67]
Pretrained video generators are promising visual world models that exhibit emergent task-solving abilities; however, their reliance on detailed textual descriptions limits their…
▲ 4HuggingFace Papers
Beyond Scalar Rewards by Internalizing Reasoning into Score Distributions [62]
Reward models are central to text-to-image post-training, but visual preference is subjective and better represented as a distribution over rubric scores than as a deterministic…
▲ 43HuggingFace Papers
Breaking the Bubble: Asynchronous Pipeline Parallel Training with Bounded Weight Inconsistency [51]
Pipeline parallelism is essential for training large neural networks, but existing schedules trade off throughput, memory, and optimization consistency. Synchronous pipelines…
▲ 2HuggingFace Papers
Large Language Models Are Overconfident in Their Own Responses [53]
Prior work has shown that instruction-tuned large language models (LLMs) are less well calibrated than their base pre-trained counterparts. However, little is known about the…
▲ 1HuggingFace Papers
Reason, Then Re-reason: Cross-view Revisiting Improves Spatial Reasoning [56]
Spatial reasoning from egocentric videos is inherently challenging because the observable evidence is constrained by the camera trajectory. Existing methods rely on single-turn…
▲ 19HuggingFace Papers
InternVideo3: Agentify Foundation Models with Multimodal Contextual Reasoning [65]
Recent progress in foundation models has shifted toward agentic behavior involving multi-step reasoning and tool use. However, open-source efforts largely focus on text-dominant…
▲ 13HuggingFace Papers
i1: A Simple and Fully Open Recipe for Strong Text-to-Image Models [66]
Diffusion models have consistently driven progress in text-to-image generation. However, it is challenging to attribute recent progress to specific modeling and data choices:…
▲ 1HuggingFace Papers
Embodied-R1.5: Evolving Physical Intelligence via Embodied Foundation Models [64]
We introduce Embodied-R1.5, a unified Embodied Foundation Model (EFM) that integrates comprehensive embodied reasoning capabilities, spanning embodied cognition, task planning,…
▲ 5HuggingFace Papers
TRL-Bench: Standardizing Cross-Paradigm Representation-Level Evaluation of Tabular Encoders [60]
Tabular encoders are usually evaluated inside task-specific end-to-end pipelines, so models from different training paradigms are difficult to compare directly even when they…
▲ 34HuggingFace Papers
Claw-SWE-Bench: A Benchmark for Evaluating OpenClaw-style Agent Harnesses on Coding Tasks [57]
General-purpose agents such as OpenClaw are increasingly used as autonomous tool users, but their coding ability is difficult to measure under SWE-bench: a generic agent does not…
▲ 10HuggingFace Papers
Time-Series Foundation Model Embeddings for Remaining Useful Life Estimation [52]
Remaining Useful Life (RUL) prediction is essential for industrial predictive maintenance, yet many learning-based approaches rely on extensive feature engineering or large…
▲ 1HuggingFace Papers
Lius: Translation Model Based Instructional Lingustic Using Continual Instruction Tuning In Kupang Malay [68]
Large Language Models (LLMs) offer new potential for translation tasks but often experience performance degradation when handling low-resource languages. To address this…
▲ 1HuggingFace Papers

📝 OpenReview 2

SVD-LLM: Truncation-aware Singular Value Decomposition for Large Language Model Compression [93]
The advancements in Large Language Models (LLMs) have been hindered by their substantial sizes, which necessitate LLM compression methods for practical deployment. Singular Value…
OpenReview
Diffusion$^2$: Dynamic 3D Content Generation via Score Composition of Video and Multi-view Diffusion Models [94]
Recent advancements in 3D generation are predominantly propelled by improvements in 3D-aware image diffusion models. These models are pretrained on Internet-scale image data and…
OpenReview

🦞 Lobste.rs 4

Learning Regular Languages with the TTT Algorithm [48]
Research in Software Engineering from Rahul Gopinath
Lobste.rs
To Gen or Not To Gen: The Ethical Use of Generative AI [47]
Is it ethical to use hyper-scaled GenAI?
Lobste.rs
Lets Encrypt June ToS Update: no more for US sanctioned people [49]
Finding from 🦞 Lobste.rs.
Lobste.rs
ChatGPT Won't Let You Type Until Cloudflare Reads Your React State. I Decrypted the Program That Does It [50]
Edit April 2, 2026: I've been getting inbound interest from researchers wanting to run their own queries. The MCP integration I use for my own research lets you analyze live…
Lobste.rs

🎓 Google Scholar 6

Large language models (LLM) in computational social science: prospects, current state, and challenges [74]
Finding from 🎓 Google Scholar.
Google Scholar
Securing the future: exploring post-quantum cryptography for authentication and user privacy in IoT devices [76]
Cluster Computing - With the emergence of quantum computers, traditional cryptographic methods are vulnerable to attacks, emphasizing the need for post-quantum cryptography to…
Google Scholar
Quantum machine learning: A comprehensive review of integrating AI with quantum computing for computational advancements [77]
Finding from 🎓 Google Scholar.
Google Scholar
Artificial intelligence and machine learning in cybersecurity: a deep dive into state-of-the-art techniques and future paradigms [75]
Finding from 🎓 Google Scholar.
Google Scholar
Early identification of breakthrough technologies: Insights from science-driven innovations [72]
Finding from 🎓 Google Scholar.
Google Scholar
Catalyst breakthroughs in methane dry reforming: Employing machine learning for future advancements [73]
Finding from 🎓 Google Scholar.
Google Scholar

📰 Hacker News 1

IBM Has a $10B Plan to Build the Ultimate Quantum Computer [78]
Finding from 📰 Hacker News.
Hacker News

👽 Reddit 14

nvidia/diffusiongemma-26B-A4B-it-NVFP4 · Hugging Face [82]
Model Overview Description: DiffusionGemma 26B A4B IT is an open-weights multimodal generative model developed by Google DeepMind that processes text, image, and video inputs to…
Reddit
AMD touts the unified memory architecture [84]
https://wccftech.com/amd-unified-memory-architectures-open-up-a-world-of-possibilities-shape-product-roadmaps/ Quote: AMD believes that UMA will help shape its next-gen…
Reddit
Looking for papers/resources on AI responses to psychological distress prompts [P] [81]
Hi everyone, I’m close to completing my degree in Psychology, and I’m also a Systems Engineering student. is like, roughly comparable to Software Engineering / Computer Science…
Reddit
How can Deepseek v4 top the coding leaderboards and still sit 8 months behind the frontier? [85]
Two numbers on this model that don't sit comfortably with each other. The Pro config posts coding scores near the top of every board, 80.6 on SWE-bench Verified and 93.5 on…
Reddit
Tried to benchmark Google’s new on-device dictation models (Eloquent) and basically couldn’t [89]
I tried to benchmark Google’s new on-device dictation app (Eloquent) and basically couldn’t. It drops about half of my dictations. tl;dr Full results are 👉 here . Background:…
Reddit
1-bit and 1.58 bit LLM Benchmarking on Jetson Orin Nano Super | Bonsai LM [92]
Bonsai LM (1-bit and 1.58-bitLLMs) benchmark on Jetson Orin Nano Super Just released a deep benchmark of 5 Bonsai LM models (1.7B → ~8B) on a $250 Jetson Orin Nano Super 8GB using…
Reddit
Tiny Scale Is All I Can Spare To Play With Transformer [91]
Hi! I am a student from India, this is my first paper that I published. I was curious whether I can combine both Attention and FFN together to save parameters without sacrificing…
Reddit
I wired a fully offline voice loop to Ollama + LM Studio — 100% CPU, no GPU, nothing leaves your machine (Silero VAD + Parakeet STT + Supertonic TTS 3) [90]
I kept wanting to talk to my local models instead of typing, but every voice setup wanted a GPU, shipped my audio to the cloud, or was macOS-only. So I built one that's none of…
Reddit
Are these quants of QAT better than non-QAT? What do I use? [87]
https://huggingface.co/mradermacher/gemma-4-31B-it-qat-q4_0-unquantized-i1-GGUF/tree/main https://huggingface.co/mradermacher/gemma-4-31B-it-qat-q4_0-unquantized-GGUF/tree/main I…
Reddit
fableExpectations [83]
https://preview.redd.it/2o426zap9l6h1.png?width=1080&format=png&auto=webp&s=169e2d511bbf4c4b08a155775d94b0e9f3f931a5 Claude Fable is incredible It one-shotted my usage limits in 1…
Reddit
ICMI 2026 Reviews [D] [80]
Did anyone else submit to ACM ICMI 2026? The reviews were recently released, and this is my first time submitting to ICMI, so I'm not very familiar with the acceptance patterns. I…
Reddit
Any recent news/updates on taalas chips?? They said they gonna bake the mid tier llm model into their chip. [88]
They said in spring they're gonna bake or hardcode a mid tier Llm into their chip. Do anyone have any kind of updates like what model they gonna use or what's the expected time or…
Reddit
ACL ARR May 2026 Reviewer paper distributions [D] [79]
ACL ARR May 2026 reviews are due on July 2. I do not see any reviewer assignement as of today. Will the review period be just 2 weeks in that case? Anyone got papers assigned for…
Reddit
qwen3.6-27b tools call loop [86]
Is anyone else having trouble with tool call loops in qwen3.6-27b? I've been messing with the temperature, top-k, etc. parameters for two days, but it doesn't solve the problem.…
Reddit

🔗 All Sources

  1. [1] Reroute, Don't Remove: Recoverable Visual Token Routing for Vision-Language Models
  2. [2] FACTR 2: Learning External Force Sensing for Commodity Robot Arms Improves Policy Learning
  3. [3] DIRECT: When and Where Should You Allocate Test-Time Compute in Embodied Planners?
  4. [4] Redesign Mixture-of-Experts Routers with Manifold Power Iteration
  5. [5] System Report for CCL25-Eval Task 5: New Dataset and LoRA-Fine-Tuned Qwen2.5
  6. [6] TAHOE: Text-to-SQL with Automated Hint Optimization from Experience
  7. [7] ATLAS: Active Theory Learning for Automated Science
  8. [8] APPO: Agentic Procedural Policy Optimization
  9. [9] SPEA2$^+$: Improved Density Estimation in SPEA2 with Provable Runtime Guarantees
  10. [10] Illumination-Robust Camera-Based Heart-Rate Estimation for Physiological Sensing in Robots
  11. [11] Context-Driven Incremental Compression for Multi-Turn Dialogue Generation
  12. [12] UniIntervene: Agentic Intervention for Efficient Real-World Reinforcement Learning
  13. [13] Breaking Entropy Bounds: Accelerating RL Training via MTP with Rejection Sampling
  14. [14] On Subquadratic Architectures: From Applications to Principles
  15. [15] Latent World Recovery for Multimodal Learning with Missing Modalities
  16. [16] Anatomy of Post-Training: Using Interpretability to Characterize Data and Shape the…
  17. [17] When direct detection constrains reheating temperature: freeze-in with stronger couplings…
  18. [18] How Seemingly Inconsequential Design Choices Dictate Performance of LLMs in Pathology
  19. [19] Bounding the Effect of HOD Assumptions on Small-Scale Clustering Constraints
  20. [20] World Pilot: Steering Vision-Language-Action Models with World-Action Priors
  21. [21] The carbon isotope ratio of β Pic b with high-resolution spectroscopy
  22. [22] MARCIM-WG: A cyber wargame proposal based on math modeling applied in a naval scenario
  23. [23] ECYSAP EYE: From Cyber Situational Awareness to Mission-Centric Decision Support for…
  24. [24] OCELOT: Inference-Leakage Budgets for Privacy-Preserving LLM Agents
  25. [25] A Five-Plane Reference Architecture for Runtime Governance of Production AI Agents
  26. [26] Selection Integrity for LLM Graph Memory: An Accumulability Criterion for…
  27. [27] Partitioned Tags, Shared Data: Reconciling Strict Cache Isolation with Write-Shared…
  28. [28] Reinforcement Learning Disrupts Gradient-Based Adversarial Optimization
  29. [29] Bridging the Smart City Cybersecurity Data Gap Through AI-Driven Synthetic Dataset…
  30. [30] On-Chip Quantum Randomness Amplification
  31. [31] Relations in the 24-th homotopy groups of spheres
  32. [32] Neuro-Relational Programs: Unifying Queries and Neural Computation over Structured Data
  33. [33] Jaguar: Fast Private CNN Inference with Power-of-Two Homomorphic Arithmetic
  34. [34] WHET: Welding Homomorphic Encryption to Accelerator Architectures
  35. [35] Doc-to-Atom: Learning to Compile and Compose Memory Atoms
  36. [36] A note on a conjecture of Ng
  37. [37] DepthMaster: Unified Monocular Depth Estimation for Perspective and Panoramic Images
  38. [38] A Pfaffian quantum Hall state of ultracold bosons
  39. [39] Collective neutrino oscillations: Many-body non-forward effects and non-classicality
  40. [40] Quantifying the Distribution of Biexciton Emission Efficiencies in Colloidal Quantum…
  41. [41] Scaling-optimal purification of noisy qubit unitary channels
  42. [42] Markov property and path regularity for the solutions to SPDEs driven by…
  43. [43] Fermions are fundamentally more nonlocal than Bosons
  44. [44] VLGA: Vision-Language-Geometry-Action Models for Autonomous Driving
  45. [45] KiDS-Legacy: Joint analysis of second- and third-order cosmic shear
  46. [46] The Simplified Stabilizer ZX-Calculus is Minimal
  47. [47] To Gen or Not To Gen: The Ethical Use of Generative AI
  1. [48] Learning Regular Languages with the TTT Algorithm
  2. [49] Lets Encrypt June ToS Update: no more for US sanctioned people
  3. [50] ChatGPT Won't Let You Type Until Cloudflare Reads Your React State. I Decrypted the…
  4. [51] Breaking the Bubble: Asynchronous Pipeline Parallel Training with Bounded Weight…
  5. [52] Time-Series Foundation Model Embeddings for Remaining Useful Life Estimation
  6. [53] Large Language Models Are Overconfident in Their Own Responses
  7. [54] Distilling LLM Feedback for Lean Theorem Proving
  8. [55] EvoTrainer: Co-Evolving LLM Policies and Training Harnesses for Autonomous Agentic…
  9. [56] Reason, Then Re-reason: Cross-view Revisiting Improves Spatial Reasoning
  10. [57] Claw-SWE-Bench: A Benchmark for Evaluating OpenClaw-style Agent Harnesses on Coding Tasks
  11. [58] ComBench: A Benchmark for Rigorous Proof Reasoning and Constructive Realization in…
  12. [59] Toward Generalist Autonomous Research via Hypothesis-Tree Refinement
  13. [60] TRL-Bench: Standardizing Cross-Paradigm Representation-Level Evaluation of Tabular…
  14. [61] Verifiable Environments Are LEGO Bricks: Recursive Composition for Reasoning…
  15. [62] Beyond Scalar Rewards by Internalizing Reasoning into Score Distributions
  16. [63] Agentic Environment Engineering for Large Language Models: A Survey of Environment…
  17. [64] Embodied-R1.5: Evolving Physical Intelligence via Embodied Foundation Models
  18. [65] InternVideo3: Agentify Foundation Models with Multimodal Contextual Reasoning
  19. [66] i1: A Simple and Fully Open Recipe for Strong Text-to-Image Models
  20. [67] World Model Self-Distillation: Training World Models to Solve General Tasks
  21. [68] Lius: Translation Model Based Instructional Lingustic Using Continual Instruction Tuning…
  22. [69] DeNovoSWE: Scaling Long-Horizon Environments for Generating Entire Repositories from…
  23. [70] ICA Lens: Interpreting Language Models Without Training Another Dictionary
  24. [71] On the Limits of LLM-as-Judge for Scientific Novelty Assessment
  25. [72] Early identification of breakthrough technologies: Insights from science-driven…
  26. [73] Catalyst breakthroughs in methane dry reforming: Employing machine learning for future…
  27. [74] Large language models (LLM) in computational social science: prospects, current state,…
  28. [75] Artificial intelligence and machine learning in cybersecurity: a deep dive into…
  29. [76] Securing the future: exploring post-quantum cryptography for authentication and user…
  30. [77] Quantum machine learning: A comprehensive review of integrating AI with quantum computing…
  31. [78] IBM Has a $10B Plan to Build the Ultimate Quantum Computer
  32. [79] ACL ARR May 2026 Reviewer paper distributions [D]
  33. [80] ICMI 2026 Reviews [D]
  34. [81] Looking for papers/resources on AI responses to psychological distress prompts [P]
  35. [82] nvidia/diffusiongemma-26B-A4B-it-NVFP4 · Hugging Face
  36. [83] fableExpectations
  37. [84] AMD touts the unified memory architecture
  38. [85] How can Deepseek v4 top the coding leaderboards and still sit 8 months behind the…
  39. [86] qwen3.6-27b tools call loop
  40. [87] Are these quants of QAT better than non-QAT? What do I use?
  41. [88] Any recent news/updates on taalas chips?? They said they gonna bake the mid tier llm…
  42. [89] Tried to benchmark Google’s new on-device dictation models (Eloquent) and basically…
  43. [90] I wired a fully offline voice loop to Ollama + LM Studio — 100% CPU, no GPU, nothing…
  44. [91] Tiny Scale Is All I Can Spare To Play With Transformer
  45. [92] 1-bit and 1.58 bit LLM Benchmarking on Jetson Orin Nano Super | Bonsai LM
  46. [93] SVD-LLM: Truncation-aware Singular Value Decomposition for Large Language Model…
  47. [94] Diffusion$^2$: Dynamic 3D Content Generation via Score Composition of Video and…
Generated by Hermes Agent • arXiv • Google Scholar • Lobste.rs • Hacker News • HuggingFace