RXed AI News

AI news with a spine.
@RXed_EU

Every Research story from RXed AI News over the last 30 days — 66 items.

New reinforcement learning method uses predictive divergence masks to stabilize LLM off-policy updates, improving training efficiency. huggingface.co/papers/2607.10848
View on X
Interactive video generation now controlled by graph-based object interactions, not just pixels. arxiv.org/abs/2607.21580v1
View on X
New research enables multimodal region control for Diffusion Transformers, giving artists precise editing power. arxiv.org/abs/2607.19344v1
View on X
New AI method reduces repetitive copying in long-context reasoning by grounding steps in evidence. arxiv.org/abs/2607.19345v1
View on X
No universally superior harness exists for autonomous discovery systems like OpenEvolve and TTT-Discover. arxiv.org/abs/2607.18235v1
2 1View on X
Active human vision is a closed loop redirecting gaze through intermediate hypotheses, not single snapshots. arxiv.org/abs/2607.16165v1
View on X
Autonomous data science agents still rely on expensive trial-and-error workflows, limiting efficiency. huggingface.co/papers/2607.15901
View on X
Muon shows competitive performance in large-scale pre-training but its value for reinforcement-learning post-training is unclear. arxiv.org/abs/2607.16169v1
View on X
@DeepMind's GenCeption video generator matches state-of-the-art vision systems using far less training data. the-decoder.com/google-deepmind-argues-video-genera…
View on X
SceneBind creates omni-modal scene representations binding vision, audio, and language for realistic environments. arxiv.org/abs/2607.15265v1
View on X
SearchOS-V1 tackles agent collaboration failures as web search becomes core to information-seeking agents. arxiv.org/abs/2607.15257v1
View on X
University of Pennsylvania professor used @OpenAI's GPT-5.6 Sol to disprove a 30-year-old statistics conjecture in 90 minutes. the-decoder.com/gpt-5-6-sol-reportedly-disproves-a-…
1View on X
GigaWorld-Policy-0.5 improves robot learning with World Action Models using future scenes as supervision. huggingface.co/papers/2607.13960
View on X
New QA-driven LLM agents reduce software issue resolution errors by acquiring repository knowledge before fixing. huggingface.co/papers/2607.11111
View on X
LLM agents struggle to gauge task complexity, often overcommitting to simple problems, new research shows. arxiv.org/abs/2607.13034v1
1View on X
SpectraReward turns pretrained MLLMs into zero-shot reward models for image generation RL, requiring no new training. huggingface.co/papers/2607.11886
View on X
CtrlVTON enables precise garment placement in virtual try-ons using visual-instance-prompt segmentation. huggingface.co/papers/2607.09362
View on X
Proxy-Guided Update Signals enable modular LLM post-training with reusable domain-specific policies. huggingface.co/papers/2607.11505
View on X
New theory explains how Transformers develop inductive reasoning abilities, with implications for future AI regulation. arxiv.org/abs/2607.11875v1
1View on X
New theory explains when speculative decoding drafts are accepted to accelerate language model inference. arxiv.org/abs/2607.11881v1
View on X
Vision-language models have made 10 years of progress in visual reasoning, reducing errors in complex human interactions. arxiv.org/abs/2607.09654v1
View on X
New AI agent benchmarks test 500-day terminal tasks, revealing critical gaps in long-horizon autonomy. huggingface.co/papers/2607.08964
View on X
A proactive memory agent surfaces critical states across long-horizon AI tasks without trajectory bloat. huggingface.co/papers/2607.08716
View on X
SLORR enables efficient model compression without accuracy loss by using simple in-training low-rank regularization. arxiv.org/abs/2607.08754v1
View on X
Beijing Academy of Artificial Intelligence's Orca world model predicts abstract states without action labels, matching robotics systems. the-decoder.com/chinas-orca-world-model-matches-spe…
View on X
@OpenAI's GPT-5.6 Sol Ultra reportedly proved the 50-year-old Cycle Double Cover Conjecture in under an hour using 64 subagents. the-decoder.com/openais-gpt-5-6-sol-ultra-reportedl…
View on X
CausalDS introduces a benchmark to evaluate causal reasoning in data-science agents using LLMs. huggingface.co/papers/2607.08093
1View on X
OpenCoF uses video generation to train models in logical reasoning without human labels. arxiv.org/abs/2607.08763v1
View on X
Scientific ideas inherit and recombine like genomes, challenging current AI benchmarks. arxiv.org/abs/2607.08758v1
View on X
New research demonstrates modular pretraining enables granular access control for AI models. lesswrong.com/posts/43vKjWuH4goLwrFHA/modular-pretr…
View on X
A mathematical theory of attention is tested against trained transformers in a new series of experiments. lesswrong.com/posts/2dA7phbYZGPjhTj9q/transformers-…
View on X
LessWrong explores the surprising historical path to democracy, a system often assumed to be inevitable. lesswrong.com/posts/cLQrijiGDa9X9wHqt/how-did-we-ge…
View on X
Neel Nanda’s research finds data filtering works far worse than expected for AI model training. alignmentforum.org/posts/aTybJ6CPQrxEY8rE2/data-fil…
View on X
New AI model classifies astronomical transients without human labels using uncertainty quantification. arxiv.org/abs/2607.05393v1
View on X
New robot vision model maintains accuracy without camera calibration, solving real-world deployment challenges. arxiv.org/abs/2607.05396v1
View on X
New reinforcement learning method distills reasoning from strong models without costly retraining. arxiv.org/abs/2607.05394v1
View on X
Bentham's Bulldog publishes a case against FDT, contrasting rationalist decision theory with pragmatic predictors in game theory. alignmentforum.org/posts/SdGbWkCZgCN7EGBxM/pragmati…
View on X
New visual token pruning method compresses redundant image patches while preserving critical cues for dense VLM instructions. arxiv.org/abs/2607.02484v1
View on X
G-RRM combines neural networks with symbolic reasoning to solve larger problems than previous methods. arxiv.org/abs/2607.02491v1
View on X
AutoMem automates memory expertise in LLMs, teaching them what to encode, when to retrieve, and how to organize knowledge. huggingface.co/papers/2607.01224
View on X
DemoPSD introduces disagreement-modulated policy self-distillation to boost LLM reasoning without multiple models. arxiv.org/abs/2607.02502v1
View on X
@Google Research's WARP recovers training data portfolios from foundation models without accessing raw data. huggingface.co/papers/2607.01686
View on X
Reasoning LLMs boost speaker recognition accuracy in long-form TV dramas, critical for storyline comprehension. arxiv.org/abs/2607.02504v1
View on X
Grid-based ANN search methods are systematically characterized for high-dimensional scaling. huggingface.co/papers/2607.01283
View on X
LLM agents develop social structures and objectives in multi-agent debates based on roles and audience. arxiv.org/abs/2607.02507v1
View on X
ReContext uses recursive evidence replay to help LLMs reason over long contexts. arxiv.org/abs/2607.02509v1
View on X
Quantum-inspired fast weight programmers achieve 99.8% accuracy in traffic matrix forecasting using 95% fewer parameters. huggingface.co/papers/2606.27821
View on X
New programming paradigm 'Program-as-Weights' tackles fuzzy functions like log analysis and JSON repair using model weights directly. arxiv.org/abs/2607.02512v1
View on X
TurboServe enables efficient, economical streaming video generation for long-lived user sessions. huggingface.co/papers/2606.19271
View on X
New research improves AI learning from imperfect human demonstrations without compressed signals. arxiv.org/abs/2607.01225v1
View on X
New research decouples perception and reasoning for fine-grained visual reasoning in vision-language models. huggingface.co/papers/2607.01191
View on X
New study measures the gap between human and LLM research ideas, finding current models fall short of human creativity. arxiv.org/abs/2607.01233v1
View on X
AutoMem automates memory expertise in LLMs, teaching them what to encode, when to retrieve, and how to organize knowledge. arxiv.org/abs/2607.01224v1
View on X
@Meta's non-invasive brain-to-text AI translates brain activity into typed sentences without surgery. the-decoder.com/metas-non-invasive-brain-to-text-ai…
View on X
New paper from @danahboyd and others on avoiding accountability decoys in AI's political economy. bespacific.com/reckoning-with-the-political-economy…
View on X
New theory explains when speculative decoding drafts are accepted to accelerate language model inference. arxiv.org/abs/2606.30265v1
View on X
Ko-WideSearch benchmarks web agents on breadth-searching closed sets, not just depth-solving single answers. huggingface.co/papers/2606.27595
View on X
Information-aware KV cache compression improves long-reasoning efficiency in large language models. huggingface.co/papers/2606.26875
View on X
Multimodal web agents autonomously decompose complex GUI tasks into executable actions for human assistance. arxiv.org/abs/2606.27330v1
View on X
Language-based digital twins show promise for early detection of mild cognitive impairment in elderly patients. arxiv.org/abs/2606.27334v1
View on X