Researchers automated LLM reasoning strategy design and cut token usage by 69.5%

Researchers from Meta, Google, and universities introduced AutoTTS, a framework that automatically discovers better test-time scaling strategies for large language models. In experiments, AutoTTS reduced token consumption by up to 69.5% without sacrificing accuracy, lowering inference costs for advanced reasoning workloads.

AutoTTS points to a shift from hand-designed prompting and inference heuristics toward automated optimization of model compute. If adopted in production, tools like this could make reasoning models more economically viable by dynamically balancing accuracy, latency, and token costs.

Test-time scaling (TTS) has emerged as a proven method to improve the performance of large language models in real-world applications by giving them extra compute cycles at inference time. However, TTS strategies have historically been handcrafted, relying heavily on human intuition to dictate the rules of the model’s reasoning. To address this bottleneck, researchers from Meta, Google, and several universities have introduced AutoTTS, a framework that automatically discovers optimal TTS strategies. This automated approach allows enterprise organizations to dynamically optimize compute allocation without manually tuning heuristics. By implementing the optimal strategies discovered by AutoTTS, organizations can directly reduce the token usage and operational costs of deploying advanced reasoning models in production environments. In experimental trials, AutoTTS managed inference budgets efficiently, successfully reducing token consumption by up to 69.5% without sacrificing accuracy.The manual bottleneck in test-time scalingTest-time scaling enhances LLMs by granting them extra compute when generating answers. This extra compute allows the model to generate multiple reasoning paths or evaluate its intermediate steps before arriving at a final response. The primary challenge for designing TTS strategies is determining how to allocate this extra computation optimally. Historically, researchers have designed these strategies manually, relying on guesswork to build rigid heuristics. Engineers must hypothesize the rules and thresholds for when a model should branch out into new reasoning paths, probe deeper into an existing path, prune an unpromising branch, or stop reasoning altogether. Because this manual tuning process is constrained by human intuition, a vast amount of possible approaches remain unexplored. This often results in suboptimal trade-offs between model accuracy and computing costs.Current TTS algorithms can be mapped to a width-depth control space — "width" being the number of reasoning branches explored, "depth" being how far each develops. Self-consistency (SC) samples a fixed number of trajectories and majority-votes the answer. Adaptive-consistency (ASC) saves compute by stopping early once a confidence threshold is hit. Parallel-probe takes a more granular approach, pruning unpromising branches while deepening the rest. All three are hand-crafted, and that's the constraint AutoTTS is designed to break.While some more advanced methods employ richer structures like tree search or external verifiers, they all share one key characteristic: they are meticulously hand-crafted. This manual approach restricts the scope of strategy discovery, leaving a massive portion of the potential resource-allocation space untouched.Automating strategy discovery with AutoTTSAutoTTS reframes the way test-time scaling is optimized. Instead of treating strategy design as a human task, AutoTTS approaches it as an algorithmic search problem within a controlled environment. This framework redefines the roles of both the human engineer and the AI model. Rather than hand-crafting specific rules for when an LLM should branch, prune, or stop reasoning, the engineer's role shifts to constructing the discovery environment. The human defines the boundaries, including the control space of states and actions, optimization objectives balancing accuracy versus cost, and the specific feedback mechanisms. An explorer LLM, such as Claude Code, designs the strategy. This explorer acts as an autonomous agent that iteratively proposes TTS “controllers.” These controllers are code-defined policies or algorithms that dictate how an AI model allocates its computational budget during inference. The explorer tests and refines these controllers based on feedback until it discovers an optimal resource-allocation policy. To make this automated search computationally affordable, AutoTTS relies on an “offline replay environment.” If the explorer LLM had to invoke a base reasoning model to generate new tokens every time it tested a new strategy, the compute costs would be astronomical. Instead, it relies on thousands of reasoning trajectories pre-collected from the base LLM. These trajectories include "probe signals," which are intermediate answers that help the controller evaluate progress across different reasoning branches. During the discovery loop, the explorer agent proposes a controller and evaluates it against this offline data. The agent observes the execution traces of the proposed controller that show it allocated compute over time. By analyzing these traces, the agent can diagnose specific failure modes, such as noting if a controller pruned branches too aggressively in a specific scenario. This provides an advantage over just viewing a final result. The agent then iteratively rewrites its code to improve the accuracy-cost tradeoff. Inside the AI-designed controllerBecause the explorer agent is not constrained by human intuiti