hierarchy_emergence_forces_phi

"the resulting model, LLaDA 1.5, outperforms its SFT-only predecessor consistently and significantly across mathematical (GSM8K +4.7), code (HumanEval +3.0, MBPP +1.8), and alignment benchmarks"

"We further show that agent capabilities increase with model and data size, and successfully train a single 317M parameter agent to perform 80 tasks across multiple task domains, embodiments, and action spaces."

"systematic bias reduction using a novel AfLite algorithm that generalizes human-detectable word associations to machine-detectable embedding associations"

"Experimental results demonstrate that our method significantly outperforms existing techniques that train robot manipulation policies from large-scale videos."

"Our 7B model, Pixel Reasoner, achieves 84% on V* bench, 74% on TallyQA-Complex, and 84% on InfographicsVQA"

"jointly learning to assess diverse visual tasks yields substantial mutual benefits... achieving consistent improvements across each domain"

"Empirical evaluations across diverse benchmarks demonstrate that our approach yields robust, scalable, and stable training, achieving a 17% improvement over base models and a 15% gain over SFT models."

"By leveraging the strong language capability of ChatGPT and abundant AI models in Hugging Face"

"Hydra-MDP learns how the environment influences the planning in an end-to-end manner instead of resorting to non-differentiable post-processing"

"NeighborhoodAggregation... message passing scheme... gather and scatter operations"

"The core of our technical approach lies in a diffusion probabilistic model that plans by iteratively denoising trajectories."

"The iTransformer model achieves state-of-the-art on challenging real-world datasets, which further empowers the Transformer family with promoted performance, generalization ability across different variates, and better utilization of arbitrary lookback windows"

"Using a large suite of numerical simulations, we demonstrate that the EHT images of Sgr A* are consistent with the expected appearance of a Kerr black hole with mass ∼4 × 10^6 M⊙"

"we demonstrate that SmolLM2 outperforms other recent small LMs including Qwen2.5-1.5B and Llama3.2-1B"

"We use the trained discriminators for image classification tasks, showing competitive performance with other unsupervised algorithms"

"we propose to amplify human effort through a partially automated labeling scheme, leveraging deep learning with humans in the loop"

"We provide inaugural evidence that a unified GPT-style transformer, augmented with large-scale video generative pre-training, exhibits remarkable generalization to multi-task visual robot manipulation."

"SnapKV achieves a consistent decoding speed with a 3.6x increase in generation speed and an 8.2x enhancement in memory efficiency"

"This “epidemic” of penalizing uncertain responses can only be addressed through a socio-technical mitigation: modifying the scoring of existing benchmarks"

"Our best agent can complete 30.6% of AndroidWorld's tasks, leaving ample room for future work."

"simple depth two neural networks already have perfect finite sample expressivity as soon as the number of parameters exceeds the number of data points"

"Through a three-round fine-tuning using Safe RLHF, we demonstrate a superior ability to mitigate harmful responses while enhancing model performance"

"ReST produces a dataset by generating samples from the policy, which are then used to improve the LLM policy using offline RL algorithms. ReST is more efficient than typical online RLHF methods because the training dataset is produced offline, which allows data reuse."

"AGENT-SAFETY BENCH encompasses 349 interaction environments and 2,000 test cases, evaluating 8 categories of safety risks and covering 10 common failure modes"

"we also introduce a new benchmark, 'EmpiricalBench,' to quantify the applicability of symbolic regression algorithms in science."

"DeepSpeed-Ulysses trains 2.5x faster with 4x longer sequence length than the existing method SOTA baseline."

"ALiBi's inductive bias towards recency also leads it to outperform multiple strong position methods on the WikiText-103 benchmark."

"We develop a framework for quantum resource estimation, abstracting the layers of the stack"

"Using the same base model, Qwen2.5-32B base, as DeepSeek-R1-Zero-Qwen-32B, our implementation achieves superior performance across AIME2024, MATH500, and GPQA Diamond"

"Additionally, we introduce a new instruction-tuned embedding model, whose performance is on par with state-of-the-art, English-only models of similar sizes."