{
"$type": "site.standard.document",
"bskyPostRef": {
"cid": "bafyreib3ysmzyl7g54jaf3vap6qqgtlusshwjzc5hrcdy5fycrpw4secka",
"uri": "at://did:plc:lk3jfj3zq4k4wxnk474axylu/app.bsky.feed.post/3mfueu4u5khr2"
},
"path": "/t/banana-chocolate-having-robot-issues-need-help-from-fellow-humans/1375317#post_1",
"publishedAt": "2026-02-27T18:26:44.000Z",
"site": "https://community.openai.com",
"textContent": "I’m using `gpt-4.1` to identify food items from photos. My prompt instructs the model to visually identify foods in an image and return ingredient names\n\n**The issue:** When I send a photo of a white bowl containing **pieces of chocolate and blueberries** , the model consistently returns “blueberries” (correct) and “banana” (wrong — these are clearly chocolate pieces).\n\nThis isn’t random — it’s reproducible across multiple calls. The prompt emphasizes visual inference and says “Only identify food that is CLEARLY VISIBLE in the image.”\n\nA few things I’m wondering:\n\n 1. Is this a known issue with `gpt-4.1` and dark-colored foods?\n\n 2. Would a different model (`gpt-4o`, etc.) handle this better?\n\n 3. Any prompt engineering tips to reduce food identification hallucinations?\n\n 4. You get what you pay for?\n\n\n\n\nThank you for taking a look and any suggestions!",
"title": "Banana != chocolate - having robot issues - need help from fellow humans"
}