Three items extend quantum-error-correction theory in directions that share machinery with this site’s earlier chiral-central-charge item; a fourth is a new theoretical framework connecting quantum spin liquids to unconventional superconductivity. The first two AI/ML items extend mechanistic-interpretability work already covered on this site; the last three examine, from three different angles, how reliable the reward signals used to train language-model agents actually are.

Quantum Information & Computing

P1. Approximate quantum error correction at chiral topological edges

AuthorsYuntai Song, Zejun Liu, Zhencheng Wang, Jong Yeon Lee, Bowen Shi
VenuearXiv:2608.06258
TagsQEC Chiral topological order Entanglement entropy Chiral central charge

P2. Beyond transversality: structure of Clifford circuits for CSS codes

AuthorsVictor V. Albert
VenuearXiv:2608.05688
TagsCSS codes QEC

P3. Experimental demonstration of the Quantum Fourier Transform on up to 100 qubits using a convolutional compilation strategy

AuthorsPaul Coote, Michael J. Biercuk, Yuval Baum
VenuearXiv:2608.05435 — validated on the IBM Quantum Platform
TagsQuantum Fourier transform Quantum computing

Many-body / materials physics

P1. Superconducting and charge-ordered phases from Dirac quantum spin liquids

AuthorsAndreas Feuerpfeil, Ronny Thomale, Subir Sachdev, Pietro M. Bonetti
VenuearXiv:2608.05277
TagsQuantum spin liquid Unconventional superconductivity

ML/AI

P1. CircuitSteer: geometrically aligned multi-layer steering via sparse autoencoder circuits

AuthorsMehrshad Saadatinia, Parsa Razmara, Ardalan Aryashad, Ali Abbasi, Seyedarmin Azizi
VenuearXiv:2608.05732
TagsSparse autoencoders Activation steering Mechanistic interpretability

P2. Reasoning errors have a region and a direction in the residual-stream trajectory of LLMs

AuthorsHamed Damirchi, Ignacio Meza De la Jara, Damith Ranasinghe, Yuhang Liu, Javen Shi
VenuearXiv:2608.05660
TagsResidual stream Mechanistic interpretability

P3. OSReward: instituting standardized evaluation for cross-platform computer-use reward models

AuthorsQiushi Sun, Kanzhi Cheng, Yian Wang, et al. (23 authors; NLP Group, The University of Hong Kong, and collaborators)
VenuearXiv:2607.28609
TagsLLM-as-judge Reward modeling

P4. EnvACE: internalizing environment dynamics via world rehearsal for agentic reinforcement learning

AuthorsZishan Xu, Zhiyuan Yao, Yuxin Chen, et al. (12 authors; Tencent)
VenuearXiv:2608.06197
TagsWorld models RL

P5. RRC: unlocking generative reward models in LLM reinforcement learning via ranking-based reward construction

AuthorsChenglong Wang, Ziming Zhu, Yifu Huo, Bei Li, et al. (12 authors)
VenuearXiv:2608.06310
TagsReward modeling LLM-as-judge RL