RAG 101: One Untuned Setting Drops Answer Accuracy from 100% to 33%
dima806_dima · x · 2026-11-05
- Dima Iakubovskyi published a Medium piece arguing that RAG retrieval doesn't fix hallucination — it just moves the blame.
- Empirical hook: tweaking a single setting almost nobody tunes dropped answer accuracy from 100% to 33%, with the model itself unchanged.
- Takeaway: pipeline/configuration choices in RAG can dominate answer quality, so debugging should target retrieval setup, not just the model.
More from coding & agent
- "Why do you still use the terminal?" Dev's savage reply goes viral — cephaloform · 2026-11-06
- Skip prompt engineering: let Opus orchestrate cheap models like Haiku at scale — teortaxesTex · 2026-11-06
- Recreating Ocarina remake's fog in three.js with fluid sim, prompt included — majidmanzarpour · 2026-11-06
- The hidden cost of agentmaxxing: delegating chores to AI agents erodes taste and creativity — bendee983 · 2026-11-06
- DeepLearning.AI launches voice agent course covering three integration patterns for adding speech — DeepLearningAI · 2026-11-06
- 9B Model Trained on Bittensor Scores 43.6% on SWE-bench, Beating Qwen3.5-9B — const_reborn · 2026-11-05