3月26日，被誉为"阿里最年青P10"的千问（Qwen）大模子灵魂东说念主物林俊旸，在月初辞职风云公论渐息之际，在X平台发布长文《从"推理式念念考"到"智能样式念念考"》，系统阐发了他对AI技巧范式演进剖析。通过这篇著述，林俊旸不仅总结了畴昔，更明晰地指向了AI畴昔竞争的委果战场——一个独特单一模子比拼、关乎系统、环境与协同的智能体新时期。

著述明晰地勾画出一条AI才调进化的道路图。林俊旸将2024-2025年界说为"推理念念考"阶段，以OpenAI o1和DeepSeek-R1为代表，其中枢树立是证明了"念念考"不错行为一种可历练、可托福的一流才调。这一阶段的现实，是通过强化学习（RL）在数学、代码等可考证畛域得到笃定性反应，从而让模子"为正确而优化，而非为合理"。但是，这背后是巨大的基础法子挑战——推理RL已从轻量级微调附件，演变为需要大限制部署、高隐约考证的系统工程问题。

不外，委果的难题远不啻于此。著述第二部分潜入探讨了"念念考模式"与"指示模式"和会的实践窘境。这一分析也照射了生意现实：阿里在Qwen3尝试和会后，后续的2507版块中Instruct与Thinking版块落寞呈现，因为无数客户在批量操作中仍需要高性价比、高可控的指示举止。

著述明确提议"智能样式念念考"（Agentic Thinking）是下一代AI的中枢范式。这标志着历练中枢从模子自己转向 "模子-环境"系统。智能体念念维的中枢是"为行动而念念考"，它必须处理纯推理模子无需面临的难题：决定何时行动、调用何种器用、处理环境的不笃定反应、在失败后改良筹商、在多轮交互中保握连贯。

林俊旸以为，在推理时期，上风源于更好的RL算法和反应信号；而在智能体时期，竞争上风将建立在更优质的环境假想、更紧密的历练-劳动一体化架构、以及更强健的智能体协同工程之上。环境自己成为一等品，其沉稳性、果真性、反应丰富度和抗过拟合才调至关迫切。同期，多智能体组织架构——由决策者、畛域众人和施行子代理组成的系统——将成为中枢智能的起首。

这篇著述不错看作念是林俊旸对于技巧理念的完好阐发，将他任职期间鼓励Qwen发展的技巧形而上学系统化输出。野蛮，这亦然一份个东说念主畴昔的宣言，著述中对"智能体时期"基础法子、环境工程迫切性的强调，示意了他看好的下一个创业或商榷标的。

全文由千问Qwen翻译：

From "Reasoning" Thinking to "Agentic" Thinking

从"推理式念念考"到"智能样式念念考"

The last two years reshaped how we evaluate models and what we expect from them. OpenAI's o1 showed that "thinking" could be a first-class capability， something you train for and expose to users. DeepSeek-R1 proved that reasoning-style post-training could be reproduced and scaled outside the original labs. OpenAI described o1 as a model trained with reinforcement learning to "think before it answers." DeepSeek positioned R1 as an open reasoning model competitive with o1.

畴昔两年重塑了咱们评估模子的方式以及对模子的盼愿。OpenAI的o1证明，"念念考"不错成为一种一流的手段——一种需要专门历练并面向用户洞开的才调。DeepSeek-R1则标明，推理立场的后历练方法不仅能在原始实验室以外重现，还能兑现限制化应用。OpenAI将o1形容为一种通过强化学习历练而成的模子，它能够在回答问题前"先进行念念考"。DeepSeek则将R1定位为一款与o1相失色的洞开式推理模子。

That phase mattered. But the first half of 2025 was mostly about reasoning thinking: how to make models spend more inference-time compute， how to train them with stronger rewards， how to expose or control that extra reasoning effort. The question now is what comes next. I believe the answer is agentic thinking: thinking in order to act， while interacting with an environment， and continuously updating plans based on feedback from the world.

阿谁阶段很迫切。但2025年上半年主要聚焦于推理念念维：如何让模子在推理时消耗更多时刻。谋略，如何用更是非的奖励来历练它们，如何露出或驾驭那种额外的推理起劲。当今的问题是：接下来该怎样作念？我以为谜底是代理念念维：即念念考——为了 在与环境互动时选定行动，并把柄来自外界的反应不休更新筹商。

1. What the Rise of o1 and R1 Actually Taught Us

o1和R1的崛起现实上教训了咱们什么

The first wave of reasoning models taught us that if we want to scale reinforcement learning in language models， we need feedback signals that are deterministic， stable， and scalable. Math， code， logic， and other verifiable domains became central because rewards in these settings are much stronger than generic preference supervision. They let RL optimize for correctness rather than plausibility. Infrastructure became critical.

第一波推理模子告诉咱们，若想在讲话模子中限制化应用强化学习，咱们就需要具备笃定性、沉稳性和可推广性的反应信号。数学、代码、逻辑过头他可考证的畛域因此成为中枢，因为在这些场景中，奖励信号远比一般的偏好监督更为有劲。它们使强化学习能够专注于追求正确性，而非只是追求合感性。与此同期，基础法子也变得至关迫切。

Once a model is trained to reason through longer trajectories， RL stops being a lightweight add-on to supervised fine-tuning. It becomes a systems problem. You need rollouts at scale， high-throughput verification， stable policy updates， efficient sampling. The emergence of reasoning models was as much an infra story as a modeling story. OpenAI described o1 as a reasoning line trained with RL， and DeepSeek R1 later reinforced that direction by showing how much dedicated algorithmic and infrastructure work reasoning-based RL demands. The first big transition: from scaling pretraining to scaling post-training for reasoning.

一朝模子经过历练能够推理更长的轨迹，强化学习便不再只是监督微调的一个轻量级附加组件。它……变成 一个系统性问题。你需要大限制部署、高隐约量考证、沉稳的政策更新以及高效的采样。推理模子的出现，其背后既触及基础法子确立，也关乎建摹自己。OpenAI 将 o1 形容为一种通过强化学习历练的推理模子，而 DeepSeek R1 自后进一步印证了这一标的，展示了——些许 针对基于推理的强化学习，需要专门的算法和基础法子责任。第一次首要升沉：从扩大预历练限制转向扩大后历练限制以兑现推理才调。

2. The Real Problem Was Never Just "Merge Thinking and Instruct"

委果的问题从来不单是是"和会念念考与指示"。

At the beginning of 2025， many of us in Qwen team had an ambitious picture in mind. The ideal system would unify thinking and instruct modes. It would support adjustable reasoning effort， similar in spirit to low / medium / high reasoning settings. Better still， it would automatically infer the appropriate amount of reasoning from the prompt and context， so the model could decide when to answer immediately， when to think longer， and when to spend much more computation on a truly difficult problem.

2025岁首，咱们Qwen团队的好多成员心中王人描述了一幅唯利是图的蓝图。梦想的系统是将兑现念念维与指示模式和谐，并支握可改革的推理力度，其理念类似于低/中/高三种推理成立。更棒的是，该系统能够把柄辅导和坎坷文自动推断出稳健的推理量：模子既能即时作答，也能采纳潜入念念考，以至在面临委果毒手的问题时，参加更多谋略资源进行精致求解。

Conceptually， this was the right direction. Qwen3 was one of the clearest public attempts. It introduced "hybrid thinking modes，" supported both thinking and non-thinking behavior in one family， emphasized controllable thinking budgets， and described a four-stage post-training pipeline that explicitly included "thinking mode fusion" after long-CoT cold start and reasoning RL.

从见地上讲，这是正确的标的。Qwen3是最明晰的公开尝试之一。它引入了"搀杂念念考模式"，在一个模子家眷中同期支握念念考和非念念考举止，强调可控的念念考预算，并形容了一个明确包含"念念考模式和会"的四阶段后历练经由，该经由位于长念念维链冷启动和推理强化学习之后。

But merging is much easier to describe than to execute well. The hard part is data. When people talk about merging thinking and instruct， they often think first about model-side compatibility: can one checkpoint support both modes， can one chat template switch between them， can one serving stack expose the right toggles. The deeper issue is that the data distributions and behavioral objectives of the two modes are substantially different.

但和会比邃密施行更容易形容。逶迤的部分是数据。当东说念主们褒贬和会念念考与指示时，他们频繁首先猜想的是模子侧的兼容性：一个查验点能否同期支握两种模式，一个聊天模板能否在它们之间切换，一个劳动栈能否露出正确的切换开关。更深层的问题是，这两种模式的数据分散和举止筹商存在现实互异。

We did not get everything right when trying to balance model merging with improving the quality and diversity of post-training data. During that revision process， we also paid close attention to how users were actually engaging with thinking and instruct modes. A strong instruct model is typically rewarded for directness， brevity， formatting compliance， low latency on repetitive， high-volume enterprise tasks such as rewriting， labeling， templated support， structured extraction， and operational QA. A strong thinking model is rewarded for spending more tokens on difficult problems， maintaining coherent intermediate structure， exploring alternative paths， and preserving enough internal computation to meaningfully improve final correctness.

咱们在尝试均衡模子合并与陶冶历练后数据的质料和万般性时，斗鱼app官网版并未实足作念到白玉无瑕。在这一改良过程中，咱们还密切柔软了用户如何现实参与具备念念考与引导两种模式。在企业级任务中，举例重写、标注、模板化支握、结构化索求以及运营质料保证等访佛性高、责任量大的场景，表现顽强的引导模子频繁因其径直性、粗略性、样式合规性以及低蔓延而受到扎眼。而表现顽强的念念考模子则因在责罚难题时消耗更多标识、保握连贯的中间结构、探索多种备选旅途，并保留迷漫的里面谋略以切实陶冶最终扫尾的正确性而备受珍摄。

These two behavior profiles pull against each other. If the merged data is not carefully curated， the result is usually mediocre in both directions: the "thinking" behavior becomes noisy， bloated， or insufficiently decisive， while the "instruct" behavior becomes less crisp， less reliable， and more expensive than what commercial users actually want.

这两种举止模式相互对消。淌若对合并后的数据不加以用心筛选，最终扫尾连续两端不恭维：所谓的"念念考"型举止变得参差不齐、肥胖不胜，或枯竭迷漫的决断力；而"指示"型举止则变得不够干脆利落、可靠性裁汰，且老本高于生意用户的需求。现实上想要。

Separation remained attractive in practice. Later in 2025， after the initial hybrid framing of Qwen3， the 2507 line shipped distinct Instruct and Thinking updates， including separate 30B and 235B variants. In commercial deployment， a large number of customers still wanted high-throughput， low-cost， highly steerable instruct behavior for batch operations. For those scenarios， merging wasn't obviously a benefit. Separating the lines allowed teams to focus on solving the data and training problems of each mode more cleanly.

分离在实践中仍颇具劝诱力。2025年晚些时候，在Qwen3首先的搀杂框架之后，2507版块推出了落寞的Instruct和Thinking更新版块，其中包括分散针对30B和235B参数目的变体。在生意部署中，无数客户仍然但愿在批量操作中兑现高隐约、低老本且高度可操控的指示举止。对于这些场景，合并彰着并不具备上风。将各条线分开，能让团队更明晰地专注于责罚每种模式的数据和历练问题。

Other labs chose the opposite route. Anthropic publicly argued for an integrated model philosophy: Claude 3.7 Sonnet was introduced as a hybrid reasoning model where users could choose ordinary responses or extended thinking， and API users could set a thinking budget. Anthropic explicitly said they believed reasoning should be an integrated capability rather than a separate model. GLM-4.5 also publicly positioned itself as a hybrid reasoning model with both thinking and non-thinking modes， unifying reasoning， coding， and agent capabilities; DeepSeek later moved in a similar direction with V3.1's "Think & Non-Think" hybrid inference.

其他实验室则采纳了天差地远的旅途。Anthropic公开倡导一种集成式模子理念：Claude 3.7 Sonnet被定位为一种搀杂推理模子，用户可采纳庸俗恢复或深度念念考模式，API用户还可设定念念考预算。Anthropic明确透露，他们以为推理当当是一种集成化的才调，牛牛app而非落寞的模子。GLM-4.5相似公开将自身定位为一种搀杂推理模子，兼具念念考与非念念考两种模式，兑现了推理、编码及智能体才调的和谐；DeepSeek随后也朝着类似标的迈进，其V3.1版块推出了"念念考与非念念考"搀杂推理功能。

The key question is whether the merge is organic. If thinking and instruct are merely co-located inside one checkpoint but still behave like two awkwardly stitched personalities， the product experience remains unnatural. A truly successful merge requires a smooth spectrum of reasoning effort. The model should be able to express multiple levels of effort， and ideally choose among them adaptively. GPT-style effort control points toward this: a policy over compute， rather than a binary switch.

重要问题在于，这种和会是否是天然有机的。淌若念念维与指示只是被安置于统一个查验点内，却仍表现为两种生硬拼接的个性，那么产物的用户体验将依然显得不天然。委果告成的和会，需要兑现推理起劲的平滑衔接变化。模子应当能够抒发不同线索的推理强度，而且最佳能自合乎地在这些线索之间作念出采纳。GPT式的起劲驾驭正朝着这一标的迈进：它经受的是对谋略资源的政策性调控，而非简单的二元开关。

3. Why Anthropic's Direction Was a Useful Corrective

为什么Anthropic的想法是一种有利的纠正程序

Anthropic's public framing around Claude 3.7 and Claude 4 was restrained. They emphasized integrated reasoning， user-controlled thinking budgets， real-world tasks， coding quality， and later the ability to use tools during extended thinking. Claude 3.7 was presented as a hybrid reasoning model with controllable budgets; Claude 4 extended that by allowing reasoning to interleave with tool use， while Anthropic simultaneously emphasized coding， long-running tasks， and agent workflows as primary goals.

Anthropic围绕Claude 3.7和Claude 4的公开表述是克制的。他们慎重强调了整合推理、用户可控的念念维预算、果真寰宇任务、代码质料，以及后期在永劫刻念念考过程中使用器用的才调。Claude 3.7被定位为一种具备可控预算的搀杂推理模子；Claude 4则在此基础上进一步拓展，允许推理与器用使用相互交汇。与此同期，Anthropic还极度强调了编码、恒久运行任务以及智能体责任流行为其主要筹商。

Producing a longer reasoning trace doesn't automatically make a model more intelligent. In many cases， excessive visible reasoning signals weak allocation. If the model is trying to reason about everything in the same verbose way， it may be failing to prioritize， failing to compress， or failing to act. Anthropic's trajectory suggested a more disciplined view: thinking should be shaped by the target workload. If the target is coding， then thinking should help with codebase navigation， planning， decomposition， error recovery， and tool orchestration. If the target is agent workflows， then thinking should improve execution quality over long horizons rather than producing impressive intermediate prose.

生成更长的推理轨迹并不会自动使模子变得更智能。在许厚情况下，过多的显式推理信号反而会导致分派服从低下。淌若模子试图以相似冗长的方式对所有内容进行推理，它很可能无法合理 prioritization，无法有用压缩，也无法选定行动。东说念主类的 轨迹标明，一种更严谨的视角更为稳健：念念考应以筹商责任量为导向。淌若筹商是编写代码，那么念念考就应有助于代码库导航、决策、判辨、舛误归附以及器用编排。淌若筹商是代理责任流，那么念念考的要点应放在陶冶恒久施行质料上，而非追求令东说念主惊艳的中间效果。

This emphasis on targeted utility points toward something larger: we are moving from the era of training models to the era of training agents. We made this explicit in the Qwen3 blog， writing that "we are transitioning from an era focused on training models to one centered on training agents，" and linking future RL advances to environmental feedback for long-horizon reasoning. An agent is a system that can formulate plans， decide when to act， use tools， perceive environment feedback， revise strategy， and continue over long horizons. It is defined by closed-loop interaction with the world.

这种对筹商导向型实用性的强调，指向了一个更为浩瀚的趋势：咱们正从模子历练时期迈向智能体历练时期。咱们在Qwen3博客中明确指出："咱们正在从一个以模子历练为中枢的时期，转型为以智能体历练为中枢的时期"，并把畴昔的强化学习进展与环境反应相鸠合，以支握永劫程的推理才调。所谓智能体，是一种能够制定筹商、决定行动时机、应用器用、感知环境反应、调理政策，并在长周期内握续运行的系统。它之是以平地风雷，就在于其与外界之间造成了闭环互动。

4. What "Agentic Thinking" Really Means

"智能样式念念考"的委果含义

Agentic thinking is a different optimization target. Reasoning thinking is usually judged by the quality of internal deliberation before a final answer: can the model solve the theorem， write the proof， produce the correct code， or pass the benchmark. Agentic thinking is about whether the model can keep making progress while interacting with an environment.

"智能样式念念考"是一种不同的优化筹商。推理念念维频繁以最终谜底之前的里面筹商质料来计算：模子能否解出定理、写出证明、生成正确的代码，或通过基准测试。而"智能样式念念考"则柔软的是，模子在与环境交互的过程中能否握续取得进展。

The central question shifts from "Can the model think long enough?" to "Can the model think in a way that sustains effective action?" Agentic thinking has to handle several things that pure reasoning models can mostly avoid:

Deciding when to stop thinking and take an action

Choosing which tool to invoke and in what order

Incorporating noisy or partial observations from the environment

Revising plans after failures

Maintaining coherence across many turns and many tool calls

Agentic thinking is a model that reasons through action.

中枢问题从"模子能否念念考迷漫长的时刻？"升沉为"模子能否以保管有用行动的方式进行念念考？"。智能样式念念考必须处理几件纯推理模子大多不错幸免的事情：

决定何时住手念念考并选定行动

采纳调用哪个器用以及调用轨则

融入来自环境的噪声或部分不雅测数据

在失败后改良筹商

在屡次轮次和屡次器用调用中保握连贯性

智能样式念念考是一个通过行动进行推理的模子

5. Why Agentic RL Infrastructure Is Harder

为什么智能体强化学习基础法子更难

Once the objective shifts from solving benchmark problems to solving interactive tasks， the RL stack changes. The infrastructure used for classical reasoning RL isn't enough. In reasoning RL， you can often treat rollouts as mostly self-contained trajectories with relatively clean evaluators. In agentic RL， the policy is embedded inside a larger harness: tool servers， browsers， terminals， search engines， simulators， execution sandboxes， API layers， memory systems， and orchestration frameworks. The environment is no longer a static verifier; it's part of the training system.

一朝筹商从责罚基准问题转向责罚交互式任务，强化学习的架构便会发生变化。用于经典推理强化学习的基础法子已不及以应答这一需求。在推理强化学习中，你频繁不错将采样轨迹视为大体自成一体的旅途，并配备相对明晰的评估器。而在代理强化学习中，政策被镶嵌一个更大的框架之中：器用劳动器、浏览器、终局、搜索引擎、模拟器、施行沙箱、API层、内存系统以及编排框架。此时，环境不再只是静态的考证者；它已成为历练系统的一部分。

This creates a new systems requirement: training and inference must be more cleanly decoupled. Without that decoupling， rollout throughput collapses. Consider a coding agent that must execute generated code against a live test harness: the inference side stalls waiting for execution feedback， the training side starves for completed trajectories， and the whole pipeline operates far below the GPU utilization you would expect from classical reasoning RL. Adding tool latency， partial observability， and stateful environments amplifies these inefficiencies. The result is that experimentation slows and becomes painful long before you reach the capability levels you are targeting.

这会创建一个新的系统要求：历练与推理必须兑现更透澈的解耦。若枯竭这种解耦，模子上线的隐约量将大幅下跌。试想一下，一个编码智能体需要针对及时测试框架施行生成的代码：推理端会因恭候施行反应而停滞不前，历练端则因枯竭已完成的轨迹而堕入饥饿状况，所有这个词活水线的运行服从远低于基于经典推理的强化学习所预期的GPU利用率。淌若再叠加器用蔓延、部分可不雅测性以及有状况环境等身分，这些低效问题便会进一步加重。其扫尾是，实验进程冉冉且充满苦处，以至在你尚未达到筹商才调水平之前，就一经堕入窘境。

The environment itself also becomes a first-class research artifact. In the SFT era， we obsessed over data diversity. In the agent era， we should obsess over environment quality: stability， realism， coverage， difficulty， diversity of states， richness of feedback， exploit resistance， and scalability of rollout generation. Environment-building has started to become a real startup category rather than a side project. If the agent is being trained to operate in production-like settings， then the environment is part of the core capability stack.

环境自己也正成为一类一流的商榷器用。在SFT时期，咱们酣醉于数据的万般性；而在智能体时期，咱们则应酣醉于环境的质料：包括沉稳性、果真性、覆盖范围、难度、状况万般性、反应丰富度、抗过拟合才调以及 rollout 生成的可推广性。环境构建已开动成为一个委果的创业畛域，而不再只是是副业款式。淌若智能体正在经受历练，以合乎类似分娩环境的运行场景，那么环境便成了中枢才调栈的迫切组成部分。

6. The Next Frontier Is More Usable Thought

下一个前沿是更易用的念念维

My expectation is that agentic thinking will become the dominant form of thinking. I think it may eventually replace much of the old static-monologue version of reasoning thinking: excessively long， isolated internal traces that try to compensate for lack of interaction by emitting more and more text. Even on very difficult math or coding tasks， a genuinely advanced system should have the right to search， simulate， execute， inspect， verify， and revise. The objective is to solve problems robustly and productively.

我的预期是，智能样式念念考将成为念念考的主导样貌。我以为它可能最终取代大部分旧的静态独白式推理念念考：那种因枯竭交互而通过输出越来越多文蓝本抵偿的、过长的、孑然的里面轨迹。即使在极度逶迤的数学或编码任务上，一个委果先进的系统也应该有权进行搜索、模拟、施行、查验、考证和改良。筹商是稳健且高效地责罚问题。

The hardest challenge in training such systems is reward hacking. As soon as the model gets meaningful tool access， reward hacking becomes much more dangerous. A model with search might learn to look up answers directly during RL. A coding agent might exploit future information in a repository， misuse logs， or discover shortcuts that invalidate the task. An environment with hidden leaks can make the policy look superhuman while actually training it to cheat. This is where the agent era becomes much more delicate than the reasoning era. Better tools make the model more useful， but they also enlarge the attack surface for spurious optimization. We should expect the next serious research bottlenecks to come from environment design， evaluator robustness， anti-cheating protocols， and more principled interfaces between policy and world. Still， the direction is clear. Tool-enabled thinking is simply more useful than isolated thinking， and has a far better chance of improving real productivity.

历练这类系统时，最毒手的挑战即是奖励舞弊。一朝模子得到了挑升旨的器用探访权限，奖励舞弊便会变得更加危急。具备搜索功能的模子可能会在强化学习过程中径直查找到谜底；编码智能体则可能利用仓库中的畴昔信息、花费日记，或发现一些能舒缓绕过任务要求的捷径。淌若环境中存在袒护舛错，智能体看似表现得超凡脱俗，实则是在被历练去舞弊。正因如斯，智能体时期比推理时期更加遁入和复杂。更强健的器用让模子变得更加有用，但同期也扩大了演叨优化的抨击面。咱们应预期，下一阶段的首要商榷瓶颈将来自环境假想、评估器的鲁棒性、反舞弊机制，以及政策与寰宇之间更具原则性的接口。尽管如斯，标的毅然明确：借助器用的念念维模式远比孑然的念念考更有价值，也更有可能切实陶冶分娩力。

Agentic thinking will also mean harness engineering. The core intelligence will increasingly come from how multiple agents are organized: an orchestrator that plans and routes work， specialized agents that act like domain experts， and sub-agents that execute narrower tasks while helping control context， avoid pollution， and preserve separation between different levels of reasoning. The future is a shift from training models to training agents， and from training agents to training systems.

智能样式念念考也将意味着对工程的独霸。中枢智能将越来越多地源自于多个代理的组织方式：一位崇拜决策与调度责任的统筹者，一群充任畛域众人的专科代理，以及一群施行更具体任务、同期协助驾驭坎坷文、幸免干与并保握不同线索推理之间阻难性的子代理。畴昔，咱们将从历练模子转向历练代理，再进一步从历练代理转向历练系统。

Conclusion

结语

The first phase of the reasoning wave established something important: RL on top of language models can produce qualitatively stronger cognition when the feedback signal is reliable and the infrastructure can support it.

推理波浪的第一阶段确立了一项迫切发现：在讲话模子之上应用强化学习，当反应信号可靠且基础法子能够守旧时，可产生质料上更强健的默契才调。

The deeper transition is from reasoning thinking to agentic thinking: from thinking longer to thinking in order to act. The core object of training has shifted. It is the model-plus-environment system， or more concretely， the agent and the harness around it. That changes what research artifacts matter most: model architecture and training data， yes， but also environment design， rollout infrastructure， evaluator robustness， and the interfaces through which multiple agents coordinate. It changes what "good thinking" means: the most useful trace for sustaining action under real-world constraints， rather than the longest or most visible one.

深线索的升沉是从推理式念念维转向行动式念念维：从更永劫刻的念念考，升沉为为了选定行动而进行的有序念念考。培训的中枢对象也随之发生了变化——如今，柔软的焦点已不再是单纯的模子自己，而是"模子+环境"这一系统，更具体地说，是智能体过头周围的生态系统。这使得哪些商榷效果最为重要也发生了改变：天然，模子架构和历练数据依然至关迫切；但与此同期，环境假想、部署基础法子、评估器的稳健性，以及多个智能体之间协同互动所依赖的各类接口，也王人变得相似迫切。这也再行界说了"邃密念念考"的含义：在现实寰宇的逼迫条目下，最能握续鼓励行动的有用轨迹，而非单纯追求最长或最显眼的轨迹。

It also changes where the competitive edge will come from. In the reasoning era， the edge came from better RL algorithms， stronger feedback signals， and more scalable training pipelines. In the agentic era， the edge will come from better environments， tighter train-serve integration， stronger harness engineering， and the ability to close the loop between a model's decisions and the consequences those decisions produce.

它也改变了竞争上风的起首。在推理时期，上风来自更优秀的强化学习算法、更强的反应信号以及更高的可推广性。历练活水线。在智能体时期，上风将来自更优质的环境、更紧密的历练与劳动一体化、更强健的模子逼迫工程牛牛app，以及兑现模子决策与其所产生后果之间闭环的才调。