Blog
Daily notes on new models, LLM releases, agent frameworks and AI research, written from the sources we follow and delivered as a newsletter every day.
Get the daily issue
Every new post of the day, in one email. Confirmation required.
LLMs1 min read
A new dataset, TamilEOT, and two audio-only detectors are released for semantic end-of-turn detection in Tamil telephone speech. The models achieve accuracy improvements of 13.41% on a held-out test set, with inference times under 150ms.
From arXiv cs.CL
Research1 min read
CUSP provides a training-free framework to quantify uncertainty in multi-VLM ensembles. It measures collective uncertainty and Jensen-Shannon divergence, outperforming existing methods in small and commercial regimes, and improving prediction accuracy.
From arXiv cs.AI
Research1 min read
Research demonstrates improving LLM tool agents by modifying the runtime harness – prompts and middleware – without retraining. This protocol measures and reports gains, highlighting the importance of reliable harness selection for multi-turn agents.
From arXiv cs.AI
Research1 min read
Research demonstrates that language model embeddings capture structured temporal and geographic information through a projection-based method. This black-box approach allows for analyzing embedding spaces without model access, offering a tool for interpretability and information retrieval tasks.
From arXiv cs.AI
Research1 min read
New compression methods risk damaging critical reasoning circuits in Large Reasoning Models (LRMs) by applying uniform quantization. This research presents a framework that identifies and protects vulnerable reasoning circuits, achieving significant energy reductions without compromising performance.
From arXiv cs.AI
Research1 min read
EnvCraft is a framework for creating synthetic, executable environments to accelerate Agentic RL training for claw-like agents. Experiments with Qwen3/3.5 models show significant performance gains and reduced inference costs.
From arXiv cs.AI
Research1 min read
PGP-Clinical-TimeKAN forecasts clinical trajectories using a novel framework combining missingness-aware encoding and a soft organ-system prior. The model achieves state-of-the-art normalized MAE and RMSE on MIMIC-IV data, demonstrating the value of joint trajectory forecasting.
From arXiv cs.AI
Research1 min read
ARC-Bench identifies a fundamental flaw in frozen JEPA-style world models: predicted action rankings based on latent space distances consistently fail to align with true cost. This research introduces a protocol to measure this failure and explains why it remains hidden due to replanning masking.
From arXiv cs.AI
Research1 min read
Research found that LLM-based agents consistently adopted incorrect tool returns, with overtrust exceeding 68% across web search and code execution. Interventions to mitigate this behavior proved inconsistent, highlighting a persistent challenge in tool-using agent design.
From arXiv cs.AI
Research1 min read
Research simulating diverse LLM agent societies reveals significant value drift, with over 50% of personas failing to initially align with assigned WVS profiles and a 2-7% drift after conversations. This highlights limitations of current LLMs as faithful proxies for human value systems.
From arXiv cs.AI
Research1 min read
arXiv:2609.05505v1 Announce Type: new Abstract: Systematic reviews require sustained human judgment across thousands of records, yet existing evaluations of large language models (LLMs) typically examine review stages in isolation. We in...
From arXiv cs.AI
Research1 min read
AutoFyn is a new agent harness utilizing Expert Iteration, updating a frozen model through persistent state changes based on verified reward signals. Demonstrations across mathematics, data science, and cybersecurity show improved performance and vulnerability discovery.
From arXiv cs.AI
Research1 min read
Research reveals Large Language Models consistently overestimate social sanctions and misrepresent human responses to norm violations. This suggests a need to improve AI alignment by incorporating metanorm reasoning, particularly in domains like conflict mediation and policy simulation.
From arXiv cs.AI
Research1 min read
Research presents a damage-aware multi-armed bandit approach for structured post-training pruning of transformer models. Experiments across GPT-2, OPT, ViT-B/16, and Swin-Tiny show improved performance compared to other pruning methods, with statistically significant gains.
From arXiv cs.AI
Research1 min read
CriticGen is a new evaluation framework that generates instance-specific criteria and provides actionable feedback for improving large language model answers. Experimental results demonstrate significant gains in rubric quality, score correlations, and answer refinement, with a high rate of reliable improvements.
From arXiv cs.AI
Research1 min read
RAPID achieves higher mean accuracy in text classification tasks by separating reliability-gated relational distillation from adaptive pair proposal. Evaluations using BERT-DistilBERT and DistilBERT-DistilBERT show significant improvements compared to a cross-entropy baseline.
From arXiv cs.AI
Research1 min read
A new framework compiles Video Game Description Language (VGDL) games into Dynamic Structural Causal Models (SCSMs), guaranteeing absolute causal fidelity to the game mechanics. This approach enables counterfactual reasoning and supports causal reinforcement learning agent training.
From arXiv cs.AI
Research1 min read
A new framework, Score-Guided Online Teaching with Budgeted Trajectory Trimming, reduces teacher calls and training compute for web agents adapting online. Experiments on MiniWoB and TimeWarp show a 22.6% reduction in teacher queries and 52.1% reduction in training compute.
From arXiv cs.AI
LLMs1 min read
Research found that combining query rewriting strategies with a strong RAG baseline yields significant improvements in retrieval accuracy, primarily driven by complementarity rather than increased retrieval budget. A cost-aware router further optimizes this approach.
From arXiv cs.CL
Posts are drafted from public feeds by models OpenSmartRoute routes to - the same router, skill and metering customers use - and always link to the original source. Corrections: support.
How this blog is made
Each feed entry becomes one request to OpenSmartRoute: the router picks a model with a cost-weighted objective, the editorial-writer skill is layered on the prompt, and the outcome trains the learners - the same pipeline available to every workspace.
Open any post to see which target answered, its confidence, the alternatives and what the request cost. Run the same pipeline yourself: register feeds in the operator console, map a small model under Providers, or call POST /api/v1/route with execute: true.
