ReasoningBank: Enabling agents to learn from experience (April 2026) Google Cloud research scientists Jun Yan and Chen-Yu Lee introduced ReasoningBank, an agent memory framework that distills generalizable reasoning strategies from both successful and failed task experiences so agents can keep learning after deployment. In the ICLR paper "ReasoningBank: Scaling Agent Self-Evolving with Reasoning Memory," the framework runs a closed loop of retrieval, extraction and consolidation, using an LLM-as-a-judge to self-assess trajectories and turn failures into preventative lessons; Google reports higher success rates and fewer task steps than baselines on web browsing and software engineering benchmarks. The code is published at github.com/google-research/reasoning-bank. April 21, 2026 Jun Yan and Chen-Yu Lee, Research Scientists, Google Cloud ReasoningBank is a novel agent memory framework that uses successful and failed experiences to distill generalizable reasoning strategies, enabling an agent to continuously learn from experience after deployment. Agents are becoming increasingly crucial in tackling complex real-world tasks, ranging from general web navigation to assisting with extensive software engineering codebases. However, as these agents transition into persistent, long-running roles in the real world, they face a critical limitation: they struggle to analyze and learn from successful and failed experiences after deployment. Agents approaching each new task without a memory mechanism will repeatedly make the same strategic errors and discard valuable insights. To address this, various forms of agent memory have been introduced to store information about past interactions for reuse. However, existing methods generally focus on saving exhaustive records of every action taken — such as the trajectory memory used in Synapse https://arxiv.org/abs/2306.07863 — or only documenting workflows summarized from successful attempts, as seen in Agent Workflow Memory https://arxiv.org/abs/2409.07429 . These approaches have two fundamental drawbacks: first, by recording detailed actions instead of tactical foresight, they fail to distill higher-level, transferable reasoning patterns; second, by over-emphasizing successful experiences, they miss out on a primary source of learning — their own failures. To bridge this gap, in our ICLR paper, " ReasoningBank: Scaling Agent Self-Evolving with Reasoning Memory https://arxiv.org/abs/2509.25140 ", we introduce a novel agent memory framework github https://github.com/google-research/reasoning-bank that distills useful insights from both successful and failed experiences for test-time self-evolution. When evaluated on web browsing and software engineering benchmarks, ReasoningBank enhances both agent effectiveness higher success rates and efficiency fewer task steps compared to baseline approaches. ReasoningBank distills global reasoning patterns into high-level, structured memories. Each structured memory item contains the following: The memory workflow operates in a continuous, closed loop of retrieval, extraction, and consolidation. Before taking action, the agent draws upon the ReasoningBank to gather relevant memories into its context. It then interacts with the environment and uses an LLM-as-a-judge https://arxiv.org/abs/2306.05685 to self-assess the resulting trajectory and extracts success insights or failure reflection. Notably, this self-judgement does not need to be perfectly accurate, as we find ReasoningBank to be quite robust against judgment noise. During extraction, the agent distills workflows and generalizable insights from the trajectory into new memories. For simplicity, we directly append these to the ReasoningBank, leaving more sophisticated consolidation strategies for future work. Crucially, unlike existing workflow memory strategies https://arxiv.org/abs/2409.07429 that only focus on successful runs, ReasoningBank actively analyzes failed experiences to source counterfactual signals and pitfalls. By distilling these mistakes into preventative lessons, ReasoningBank builds powerful strategic guardrails. For example, instead of merely learning a procedural rule like "click the 'Load More' button”, the agent might learn from a past failure to "always verify the current page identifier first to avoid infinite scroll traps before attempting to load more results”. Test-time scaling https://arxiv.org/abs/2408.03314 TTS — scaling compute at inference time — has shown immense effectiveness in reasoning domains like math https://arxiv.org/abs/2501.19393 and competitive programming https://arxiv.org/abs/2502.14382 . However, in agentic environments, existing TTS methods often discard the exploration trajectory and treat the final answer as the only useful outcome. This overlooked exploration is actually a rich data source that could accelerate an agent's ability to learn from experience over time. We bridge this gap by explicitly linking memory with scaling through memory-aware test-time scaling MaTTS . By using ReasoningBank as a powerful experience learner, MaTTS distills extensive exploration into high-quality memories via contrastive and refinement signals. We demonstrate the power of MaTTS functions through two distinct forms of scaling: MaTTS establishes a strong synergy: high-quality memory from ReasoningBank steers the scaled exploration towards more promising strategies, and in return, the scaled interactions generate significantly richer learning signals that feed back into an even smarter ReasoningBank to help the agent. We evaluated ReasoningBank across challenging benchmarks covering dynamic environments. Using the ReAct https://arxiv.org/abs/2210.03629 prompting strategy as the foundation for all agents, we compared ReasoningBank against three memory configurations: a memory-free baseline Vanilla ReAct , Synapse https://arxiv.org/abs/2306.07863 Trajectory Memory and AWM https://arxiv.org/abs/2409.07429 Workflow Memory . From our main evaluation results with Gemini-2.5-Flash https://docs.cloud.google.com/vertex-ai/generative-ai/docs/models/gemini/2-5-flash on WebArena https://arxiv.org/abs/2307.13854 and SWE-Bench-Verified https://openai.com/index/introducing-swe-bench-verified/ , we have the following key observations: Importantly, during evaluation, we observed the emergence of strategic maturity. In a web-browsing example, the agent's initial curated rules resembled simple procedural checklists e.g., "Look for page links" . As the agent persisted through more problem sets, these memories were incorporated during execution. Building upon existing knowledge, the agent distilled new trajectories into more advanced memories. Over time, simple checklists evolved into memories with compositional, preventative logic structures e.g., "Cross-reference tasks continuously with active page filters to ensure retrieved datasets aren't paginated prematurely" . See the paper https://arxiv.org/abs/2509.25140 for more details. ReasoningBank provides a powerful framework for enabling LLMs to learn from experiences and evolve into continuous learners during test-time. We believe memory-driven experience scaling represents a crucial new frontier for agent scaling. We are excited to share this with the broader research community. This research was conducted by Siru Ouyang, Jun Yan, I-Hung Hsu, Yanfei Chen, Ke Jiang, Zifeng Wang, Rujun Han, Long T. Le, Samira Daruki, Xiangru Tang, Vishy Tirumalashetty, George Lee, Mahsan Rofouei, Hangfei Lin, Jiawei Han, Chen-Yu Lee, and Tomas Pfister.