Reinforcement Learning

TL;DR for operators Every GUI automation project has a familiar failure mode: the agent gets almost there, makes one bad click, and the training system treats the whole episode as garbage. That is tidy for spreadsheets and absurd for learning. ProgRM addresses that absurdity by replacing final-only success/failure rewards with step-level estimates of task progress.1 Instead of asking only, “Did the agent finish?”, it asks, “How much closer is the agent now than it was one step ago?” The reward is the change in estimated progress. A search that reaches the right article but fails to bookmark it is no longer equivalent to an agent staring at the home screen and scrolling like a caffeinated intern. ...

TL;DR for operators HVAC optimisation is not really about “setting the right temperature”. That is the version suitable for brochure copy and mildly insulting procurement decks. The harder problem is deciding when comfort, occupancy, outdoor conditions, and electricity prices should overrule one another. The paper behind this article proposes a human-in-the-loop reinforcement learning controller for HVAC systems.1 Its main idea is simple enough to be useful: when occupants override the system, that feedback should not merely fix the current moment. It should also teach the controller what went wrong, so future decisions require fewer overrides. ...

TL;DR for operators A robot that works alone is already expensive, brittle, and rude to your maintenance budget. A group of robots that must work together adds a different class of difficulty: timing, communication, role allocation, shared perception, physical interference, changing team composition, and the occasional human wandering into the scene with a clipboard. ...

TL;DR for operators The paper is about air-purifying booth placement in Delhi, but the useful business lesson is broader: optimisation is rarely about chasing the loudest metric. In the study, a greedy strategy that targets the highest-AQI cells achieves the highest overall AQI improvement, at 25.76%. The PPO-based strategy is slightly lower on that headline number, at 25.39%, but much stronger on population impact and traffic impact, with zero green-space violations. ...

TL;DR for operators GUI automation agents do not usually fail because clicking is hard. They fail because almost everything they could click is irrelevant. The CoGA paper proposes a pragmatic way to reduce that waste: use a vision-language model before reinforcement learning begins to generate executable code that identifies which GUI actions are currently affordable, then use that code as an action mask during RL training and inference.1 The VLM is not the agent. It is more like an expensive consultant brought in once to write a rule-based narrowing function. After that, a reinforcement learning agent still learns the policy. ...

TL;DR for operators A smart agent can still be a bad decision-maker. That is the useful, slightly annoying lesson from LLMs are Greedy Agents: Effects of RL Fine-tuning on Decision-Making Abilities.1 The paper studies Gemma2 models acting in simple decision environments and finds that they often fail not because they cannot describe the right strategy, but because they do not reliably execute it. ...

TL;DR for operators Finance AI is moving past the parlour trick stage. The interesting question is no longer whether a large language model can read a financial headline and produce a plausible answer. Of course it can. The useful question is whether that answer can be converted into a measurable, governed, risk-aware decision process without accidentally building a very expensive rumour amplifier. ...

TL;DR for operators Most enterprise AI failures do not come from the model being “too small”. They come from the system around the model being too vague. A model gives an answer. The workflow accepts it. Nobody knows whether the reasoning path was valid, whether the data path was stale, whether the tool should have been called, or whether the whole process should be redesigned after repeated mistakes. Then someone asks why the AI confidently did something expensive. Excellent. We have automated the intern, but forgot to hire the supervisor. ...

TL;DR for operators Search-contempt is not “AI plays worse so it learns more”. That would be the lazy interpretation, and business strategy already has enough lazy interpretations wearing expensive shoes. The paper introduces a hybrid MCTS method for AlphaZero-like self-play systems. It behaves like standard PUCT search for the player to move, but at opponent nodes it eventually freezes the opponent’s visit distribution after a threshold, $N_{scl}$, and samples from that frozen distribution rather than constantly updating it toward stronger play.1 The effect is subtle but important: the system stops assuming the opponent will always improve its response with more search. ...

TL;DR for operators A Gomoku-playing LLM is not going to walk into your Monday strategy meeting and outperform the CFO. The interesting part is more useful than that. Hui Wang’s LLM-Gomoku paper shows a language model being turned into a strategic game player by surrounding it with structure: board-state representation, explicit rules, strategy prompts, local position scoring, self-play, reinforcement learning, state-action-reward storage, and visualisation.1 That is the part worth stealing. Not the board game. Not the romance of “AI intuition.” The machinery. ...