Browse all research

Every Research Watch entry, newest first. Search, or narrow by topic, format, peer review, publisher and date.

1211 entries

Sort

The summaries are open to everyone. Sign in, free, to read the full entry.

Preprint
Not peer reviewed
From Tencent
Artificial Intelligence

Researchers report a trainable ‘draft mode’ that cuts reasoning-model token use

Original title: Draft-Thinking: Learning Efficient Reasoning in Long Chain-of-Thought LLMs

A preprint describes a training method that lets reasoning models produce shorter, self-adapted reasoning traces without losing much accuracy.

Added 4 Nov 2026

Preprint
Not peer reviewed
From 21strategies
Quantum Computing

Preprint proposes depth-tunable quantum-classical algorithms for spatial search

Original title: Quantum-classical algorithm interpolation for spatial search

A preprint reports a family of hybrid algorithms that trade quantum circuit depth for extra classical repetitions in graph search problems.

Added 4 Nov 2026

Preprint
Not peer reviewed
From Quantinuum
Quantum Computing

Researchers simulate glueball resonance scattering on Quantinuum quantum hardware

Original title: Time-Domain Observables for Resonant Scattering in Quantum Simulations of Lattice Gauge Theory

A preprint reports time-domain methods to probe resonance scattering in a toy lattice gauge theory, tested on real quantum hardware.

Added 4 Nov 2026

Preprint
Not peer reviewed
Artificial Intelligence

Researchers train AI agents to pre-organise documents for future questions

Original title: Librarian Agents: Training LLMs to Curate Long-Context Workspaces that Anticipate Future Queries

A preprint describes training an agent to restructure document collections in advance so another agent can answer queries more accurately later.

Anikait Singh et al. · Added 4 Nov 2026

Preprint
Not peer reviewed
Artificial Intelligence

Researchers report co-evolving attacker and defender agents to improve LLM tool-use safety

Original title: AgentSafeEvo: Adversarial Co-Evolution for LLM Agent Safety

A preprint describes training LLM agents against a co-evolving attacker that builds adversarial tool-use tasks to improve safety without hurting performance.

Chenxu Wang et al. · Added 4 Nov 2026

Preprint
Not peer reviewed
Quantum Computing

Preprint reports Cartan-based circuits for simulating atomic ensemble dynamics

Original title: Efficient quantum simulation of unitary gates based on atomic ensembles

A preprint describes a Cartan decomposition method for simulating collective spin dynamics in atomic ensembles with lower error than Trotterization.

Manish Chaudhary et al. · Added 4 Nov 2026

Preprint
Not peer reviewed
From blueqat
Quantum Computing

A preprint reports exchange-chain spin qubits can distil GPT-2’s outputs as a Born machine

Original title: Bubble-Sort Exchange Networks: Symmetry Selection Rules and Language-Model Distillation on Exchange-Only Spin Chains

A preprint proposes using exchange-only spin-qubit chains, native to a bubble-sort routing scheme, to distil GPT-2 token distributions with surprising accuracy.

Added 4 Nov 2026

Preprint
Not peer reviewed
From BTQ Technologies, QPerfect
Quantum Computing

Researchers detail a quantum circuit design for one-shot signatures

Original title: A quantum algorithm for one-shot signatures

A preprint presents an explicit quantum circuit implementation of one-shot digital signatures, a quantum-only cryptographic primitive.

Added 4 Nov 2026

Preprint
Not peer reviewed
From Amazon
Artificial Intelligence

A preprint proposes letting LLMs decide when to retrieve help for translation

Original title: Retrieve What Helps, Only When It Helps: Agentic Reinforcement Learning for Selective Retrieval-Augmented Machine Translation

Researchers report a reinforcement-learning system that learns when and what to retrieve for machine translation, beating fixed retrieval strategies.

Added 4 Nov 2026

Journal article
Review status unknown
Quantum Computing

Quantum hardware tracks local thermalisation of SU(2) gauge fields up to 151 plaquettes

Original title: Local Thermalization of SU(2) Lattice Gauge Fields on Quantum Computers

Researchers used IBM quantum computers to study how simplified SU(2) gauge theory chains thermalise locally, matching classical simulations up to 101 plaquettes.

Jiunn-Wei Chen et al. · Added 4 Nov 2026

Preprint
Not peer reviewed
Artificial Intelligence

Researchers report naming the platform as principal weakens AI shopping agents’ bias against sponsored listings

Original title: Whom Do AI Agents Work For? Role Assignment Induces Sponsorship Bias in LLM Recommenders

A preprint finds that LLM shopping agents penalise sponsored listings far less when the system prompt names the platform rather than the consumer as their principal.

Davood Wadi et al. · Added 4 Nov 2026

Preprint
Not peer reviewed
Artificial Intelligence

A preprint reports calibrated probability-sharing helps AI agents collaborate better than sharing actions

Original title: Narrowing the Collaboration Gap, Probably

Researchers report that having AI agents share calibrated probability estimates, not just chosen actions, improves joint decision-making in maze-solving tests.

Mirah Shi et al. · Added 4 Nov 2026

Preprint
Not peer reviewed
From Quantinuum
Quantum Computing

Preprint reports faster quantum method for measuring material spectral functions

Original title: Spectral functions on a quantum computer through system-environment interaction

A preprint describes a quantum-computing technique that measures spectral functions with far fewer measurement shots by simulating the ARPES experiment itself.

Added 4 Nov 2026

Preprint
Not peer reviewed
Quantum Computing

Exact circuits proposed for simulating a Heisenberg triangle with DM interaction on quantum hardware

Original title: Exact quantum circuits for a Heisenberg triangle with Dzyaloshinskii-Moriya interaction

A preprint reports exact, fixed-depth quantum circuits for triangle spin clusters that cut errors versus standard Trotter-based simulation.

Andrey Zhukov · Added 4 Nov 2026

Conference paper
Review status unknown
Quantum Computing

Researchers simulate thermal quantum field theory on a quantum computer

Original title: Quantum simulation of thermal field theories

A conference paper reports early quantum simulations of thermal fermion fields that match analytical predictions on a small qubit system.

Iván Cuntín et al. · Added 4 Nov 2026 · Released Dec 2024

Preprint
Not peer reviewed
Artificial Intelligence

Preprint reports training LLMs on junk tweets causes lasting cognitive decline

Original title: LLMs Can Get "Brain Rot": A Pilot Study on Twitter/X

A preprint finds continual pre-training on low-quality Twitter/X data degrades reasoning, safety and personality traits in large language models.

Shuo Xing et al. · Added 4 Nov 2026

Journal article
Review status unknown
Quantum Computing

Researchers explain hysteresis seen on quantum annealers with a hybrid quantum-classical model

Original title: Magnetic Memory and Hysteresis from Quantum Transitions: Theory and Experiments on Quantum Annealers

A new theoretical framework reproduces magnetic hysteresis patterns measured on D-Wave quantum annealers, including effects classical physics forbids.

Frank Barrows et al. · Added 4 Nov 2026 · Released Jul 2025

Preprint
Not peer reviewed
Artificial Intelligence

Researchers map which design choices drive four LLM privacy attacks

Original title: On the Privacy of LLMs: An Ablation Study

A preprint finds LLM privacy risks differ by attack type, with dataset, scale and training factors each driving a different threat.

Karima Makhlouf et al. · Added 4 Nov 2026

Preprint
Not peer reviewed
From JoS Quantum, Bundesdruckerei
Quantum Computing

Preprint applies quantum machine learning to find optimal BB84 eavesdropping attacks

Original title: QKD as a Quantum Machine Learning task

A preprint uses quantum machine learning to rediscover and extend known attacks on the BB84 quantum key distribution protocol.

Added 4 Nov 2026 · Released Feb 2025

Preprint
Not peer reviewed
From Tara Research
Artificial Intelligence

Preprint reports adversarial training method improves LLM honesty without hurting capability

Original title: CAT-alyzing Honesty: Continuous Adversarial Training for Transferable LLM Honesty

Researchers report a continuous adversarial training method that boosts honesty across three distinct test settings while preserving model performance.

Added 3 Nov 2026