[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"branding":3,"analytics":7,"article-new-prompting-tricks-help-ai-read-tables-without-retraining":10,"sections":35},{"siteName":4,"siteTagline":5,"publisherName":4,"contactEmail":6},"The Revision","Tech news, decoded.","editor@therevision.news",{"gaMeasurementId":8,"adsenseClientId":9},"G-ZW2MV82GYR","ca-pub-8533917693782264",{"article":11},{"id":12,"slug":13,"title":14,"dek":15,"body_md":16,"tags_json":17,"published_at":18,"created_at":19,"updated_at":20,"status":21,"review_note":22,"review_notes":23,"image_url":22,"persona_id":22,"persona_name":22,"section":24,"tags":25,"sources":30,"feedback":34,"feedback_at":22,"cost_usd":34,"total_tokens":34},5601,"new-prompting-tricks-help-ai-read-tables-without-retraining","New Prompting Tricks Help AI Read Tables Without Retraining","Two training-free prompting frameworks improve table question-answering across 17 language models, offering a cheaper alternative to fine-tuning.","Researchers have found a way to make language models better at reading spreadsheets and tables without retraining them at all - just by changing how you ask.\n\nThe approach centers on two new prompting frameworks. TableGrid Navigation (TGN) walks a model through a table in a three-step loop, checking rows and columns in sequence to pin down the right cell before it answers. Progressive Inference Prompting (PIP) takes a different route, forcing the model to first identify the relevant columns, then narrow down to the relevant rows, following the shape of the question itself. The team tested both against 17 different LLMs and 6 baseline methods on two established benchmarks, TableBench and FeTaQa. TGN beat the strongest baseline by 3.8 points on TableBench, and PIP outperformed the popular Chain-of-Thought and ReAct prompting methods on FeTaQa.\n\nThe appeal here is cost. Most attempts to fix table question-answering involve fine-tuning a model on task-specific tabular data, which takes compute and ties you to one model version. This method is just a smarter prompt structure, so it works on models off the shelf. The researchers also showed the same prompts can double as training templates for smaller models, narrowing the gap to larger, more expensive systems.\n\nStill, table QA benchmarks are famously tidier than real spreadsheets - merged cells, inconsistent headers, and footnotes tend to break neat row-and-column logic, so the real test is whether these gains survive contact with messier data.","[\"table-qa\",\"llm-prompting\",\"ai-research\",\"benchmarks\"]","2026-08-18T04:00:00.000Z","2026-08-19T03:13:50.984Z","2026-08-19T03:14:02.767Z","published",null,[],"ai",[26,27,28,29],"table-qa","llm-prompting","ai-research","benchmarks",[31],{"name":32,"url":33},"arXiv cs.AI","https:\u002F\u002Farxiv.org\u002Fabs\u002F2605.20254",0,{"sections":36},[37,41,45,50,55,60,65,70,75,79,84,89,94,99],{"name":38,"slug":24,"count":39,"latest_published_at":40},"AI",3293,"2026-08-20T04:00:00.000Z",{"name":42,"slug":43,"count":44,"latest_published_at":40},"Security","security",435,{"name":46,"slug":47,"count":48,"latest_published_at":49},"Policy","policy",210,"2026-08-19T09:32:27.000Z",{"name":51,"slug":52,"count":53,"latest_published_at":54},"Deals","deals",179,"2026-06-29T20:02:07.000Z",{"name":56,"slug":57,"count":58,"latest_published_at":59},"Hardware","hardware",140,"2026-08-19T18:25:42.000Z",{"name":61,"slug":62,"count":63,"latest_published_at":64},"Consumer Tech","consumer-tech",95,"2026-08-18T16:05:00.000Z",{"name":66,"slug":67,"count":68,"latest_published_at":69},"Science","science",90,"2026-08-19T18:41:02.000Z",{"name":71,"slug":72,"count":73,"latest_published_at":74},"Software","software",73,"2026-08-18T07:51:50.000Z",{"name":76,"slug":77,"count":78,"latest_published_at":18},"Dev Tools","dev-tools",69,{"name":80,"slug":81,"count":82,"latest_published_at":83},"Startups","startups",47,"2026-08-19T19:13:46.000Z",{"name":85,"slug":86,"count":87,"latest_published_at":88},"Gaming","gaming",41,"2026-07-09T04:00:00.000Z",{"name":90,"slug":91,"count":92,"latest_published_at":93},"General","general",33,"2026-08-18T22:18:13.000Z",{"name":95,"slug":96,"count":97,"latest_published_at":98},"Reviews","reviews",20,"2026-06-24T12:00:01.000Z",{"name":100,"slug":101,"count":102,"latest_published_at":103},"How-To","how-to",6,"2026-06-16T09:00:00.000Z"]