{"arxiv_id":"2605.24598","incoming":{"papers":1,"sentences":1,"groups":[],"unlabelled_sentences":1},"outgoing":{"papers":12,"sentences":13,"groups":[{"kind":"method_or_tool","label":"Uses the method or tool of","papers":1,"sentences":2,"quotes":[{"arxiv_id":"2412.15115","title":"Qwen2.5 Technical Report","url":"/paper/2412.15115","section":"Experimental Settings","sentence":"We use the Qwen2.5-7B-Instruct [ref] as the on-device model and the closed-source Qwen-Max API [ref] as the strong cloud oracle, respectively (other model pairs in Appendix [ref]). is initialized from the Qwen2.5-0.5B backbone [ref] and replaces the LM head with a binary classification head, resulting in a 494M-parameter model.","intent_p":0.99,"contests_p":null}]},{"kind":"background","label":"Cites as background","papers":8,"sentences":8,"quotes":[{"arxiv_id":"2512.02556","title":"DeepSeek-V3.2: Pushing the Frontier of Open Large Language Models","url":"/paper/2512.02556","section":"Related work","sentence":"Recent work uses supervised fine-tuning [ref] and, in particular, reinforcement learning [ref] to enable LLM agents to learn directly from environment interactions rather than static prompts or handcrafted workflows, significantly enhancing long-horizon reasoning and multi-step decision-making and forming a promising paradigm for general-purpose autonomous intelligence [ref] .","intent_p":0.83,"contests_p":null},{"arxiv_id":"2507.06261","title":"Gemini 2.5: Pushing the Frontier with Advanced Reasoning, Multimodality, Long Context, and Next Generation Agentic Capabilities","url":"/paper/2507.06261","section":"Related work","sentence":"Recent work uses supervised fine-tuning [ref] and, in particular, reinforcement learning [ref] to enable LLM agents to learn directly from environment interactions rather than static prompts or handcrafted workflows, significantly enhancing long-horizon reasoning and multi-step decision-making and forming a promising paradigm for general-purpose autonomous intelligence [ref] .","intent_p":0.83,"contests_p":null},{"arxiv_id":"2602.00994","title":"Reasoning and Tool-use Compete in Agentic RL:From Quantifying Interference to Disentangled Tuning","url":"/paper/2602.00994","section":"Related work","sentence":"Recent work uses supervised fine-tuning [ref] and, in particular, reinforcement learning [ref] to enable LLM agents to learn directly from environment interactions rather than static prompts or handcrafted workflows, significantly enhancing long-horizon reasoning and multi-step decision-making and forming a promising paradigm for general-purpose autonomous intelligence [ref] .","intent_p":0.83,"contests_p":null},{"arxiv_id":"1707.06347","title":"Proximal Policy Optimization Algorithms","url":"/paper/1707.06347","section":"Preliminaries","sentence":"An episode produces a trajectory \\(τ= {(𝐬_1, 𝐚_1 ,r_1), (𝐬_2, 𝐚_2 ,r_2),…, (𝐬_T, 𝐚_T ,r_T)}\\) with an associated return \\(R(τ)\\) (e.g., verifiable, outcome-level task feedback), which is often used to optimize the LLM policy to maximize the expected return of the agent’s trajectories [ref] .","intent_p":0.87,"contests_p":null},{"arxiv_id":"2402.03300","title":"DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models","url":"/paper/2402.03300","section":"Preliminaries","sentence":"An episode produces a trajectory \\(τ= {(𝐬_1, 𝐚_1 ,r_1), (𝐬_2, 𝐚_2 ,r_2),…, (𝐬_T, 𝐚_T ,r_T)}\\) with an associated return \\(R(τ)\\) (e.g., verifiable, outcome-level task feedback), which is often used to optimize the LLM policy to maximize the expected return of the agent’s trajectories [ref] .","intent_p":0.87,"contests_p":null}]}],"unlabelled_sentences":3},"claims":[],"thresholds":{"intent_p":0.8,"supports_p":0.95,"contests_p":0.7},"method":"Citing sentences come from each paper's LaTeX source; every \\cite marks the sentence and the works it cites. Pith's classifier typed each sentence. A label is shown at intent probability 0.8 or more (0.95 for support), a contest at 0.7 or more. Measured accuracy against two blind model families is at https://pith.science/accuracy. The quoted sentences are the citing authors' own.","held_up":null,"_links":{"html":"https://pith.science/paper/2605.24598"}}