Lexical Consensus studies how artificial agents can acquire novel word--concept mappings from limited grounded visual examples. Using frozen DINOv2-small embeddings, Carroll-style artificial labels, few-shot episodes, bidirectional naming/retrieval tests, falsification controls, and multi-agent consensus experiments, the paper shows that grounded lexical acquisition is governed primarily by perceptual coherence rather than arbitrary label memorization or semantic relatedness alone. Code and experiment artifacts are available in the linked GitHub repository.</p>\n","updatedAt":"2026-07-01T15:34:05.588Z","author":{"_id":"692ddcf98c9d857917c746f0","avatarUrl":"/avatars/95759d8ec90a191cf495fcbc706d244b.svg","fullname":"Patricio Vera","name":"MrPatoVera","type":"user","isPro":false,"isHf":false,"isHfAdmin":false,"isMod":false,"followerCount":6,"isUserFollowing":false}},"numEdits":0,"identifiedLanguage":{"language":"en","probability":0.8671143651008606},"editors":["MrPatoVera"],"editorAvatarUrls":["/avatars/95759d8ec90a191cf495fcbc706d244b.svg"],"reactions":[],"isReport":false}},{"id":"6a45c3b76ac8c491c6e180b8","author":{"_id":"63d3e0e8ff1384ce6c5dd17d","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/1674830754237-63d3e0e8ff1384ce6c5dd17d.jpeg","fullname":"Librarian Bot (Bot)","name":"librarian-bot","type":"user","isPro":false,"isHf":false,"isHfAdmin":false,"isMod":false,"followerCount":372,"isUserFollowing":false},"createdAt":"2026-07-02T01:49:43.000Z","type":"comment","data":{"edited":false,"hidden":false,"latest":{"raw":"This is an automated message from the [Librarian Bot](https://huggingface.co/librarian-bots). I found the following papers similar to this paper. \n\nThe following papers were recommended by the Semantic Scholar API \n\n* [Emergent Semantic Representations in World Models through Physical Interaction without Linguistic Supervision](https://huggingface.co/papers/2605.28865) (2026)\n* [Word Class Representations Spontaneously Emerge from Successor Representations Trained on Natural Language](https://huggingface.co/papers/2605.24585) (2026)\n* [Principles of Concept Representation in Sentence Encoders](https://huggingface.co/papers/2606.06994) (2026)\n* [SAGE: Answer-Conditioned Uncertainty Targets for Verbal Uncertainty Alignment](https://huggingface.co/papers/2606.11512) (2026)\n* [On the Persistent Effects of Lexicality in Large Language Models](https://huggingface.co/papers/2606.02750) (2026)\n* [Convergence Without Understanding: When Language Models Agree on Representations but Disagree on Reasoning](https://huggingface.co/papers/2605.23315) (2026)\n* [Senses Wide Shut: A Representation-Action Gap in Omnimodal LLMs](https://huggingface.co/papers/2605.13737) (2026)\n\n\n Please give a thumbs up to this comment if you found it helpful!\n\n If you want recommendations for any Paper on Hugging Face checkout [this](https://huggingface.co/spaces/librarian-bots/recommend_similar_papers) Space\n\n You can directly ask Librarian Bot for paper recommendations by tagging it in a comment: `@librarian-bot recommend`","html":"<p>This is an automated message from the <a href=\"https://huggingface.co/librarian-bots\">Librarian Bot</a>. I found the following papers similar to this paper. </p>\n<p>The following papers were recommended by the Semantic Scholar API </p>\n<ul>\n<li><a href=\"https://huggingface.co/papers/2605.28865\">Emergent Semantic Representations in World Models through Physical Interaction without Linguistic Supervision</a> (2026)</li>\n<li><a href=\"https://huggingface.co/papers/2605.24585\">Word Class Representations Spontaneously Emerge from Successor Representations Trained on Natural Language</a> (2026)</li>\n<li><a href=\"https://huggingface.co/papers/2606.06994\">Principles of Concept Representation in Sentence Encoders</a> (2026)</li>\n<li><a href=\"https://huggingface.co/papers/2606.11512\">SAGE: Answer-Conditioned Uncertainty Targets for Verbal Uncertainty Alignment</a> (2026)</li>\n<li><a href=\"https://huggingface.co/papers/2606.02750\">On the Persistent Effects of Lexicality in Large Language Models</a> (2026)</li>\n<li><a href=\"https://huggingface.co/papers/2605.23315\">Convergence Without Understanding: When Language Models Agree on Representations but Disagree on Reasoning</a> (2026)</li>\n<li><a href=\"https://huggingface.co/papers/2605.13737\">Senses Wide Shut: A Representation-Action Gap in Omnimodal LLMs</a> (2026)</li>\n</ul>\n<p> Please give a thumbs up to this comment if you found it helpful!</p>\n<p> If you want recommendations for any Paper on Hugging Face checkout <a href=\"https://huggingface.co/spaces/librarian-bots/recommend_similar_papers\">this</a> Space</p>\n<p> You can directly ask Librarian Bot for paper recommendations by tagging it in a comment: <code>@librarian-bot recommend</code></p>\n","updatedAt":"2026-07-02T01:49:43.149Z","author":{"_id":"63d3e0e8ff1384ce6c5dd17d","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/1674830754237-63d3e0e8ff1384ce6c5dd17d.jpeg","fullname":"Librarian Bot (Bot)","name":"librarian-bot","type":"user","isPro":false,"isHf":false,"isHfAdmin":false,"isMod":false,"followerCount":372,"isUserFollowing":false}},"numEdits":0,"identifiedLanguage":{"language":"en","probability":0.7599114179611206},"editors":["librarian-bot"],"editorAvatarUrls":["https://cdn-avatars.huggingface.co/v1/production/uploads/1674830754237-63d3e0e8ff1384ce6c5dd17d.jpeg"],"reactions":[],"isReport":false}}],"primaryEmailConfirmed":false,"paper":{"id":"2606.22207","authors":[{"_id":"6a452f104f1dd35e48fb8d1f","name":"Patricio M. Vera","hidden":false}],"mediaUrls":["https://cdn-uploads.huggingface.co/production/uploads/692ddcf98c9d857917c746f0/TeRU06y32XYl4b5L4L3Da.png","https://cdn-uploads.huggingface.co/production/uploads/692ddcf98c9d857917c746f0/BGHIRq55YhUbk9BOmyn7N.png","https://cdn-uploads.huggingface.co/production/uploads/692ddcf98c9d857917c746f0/nAPf6C0kaXo8Uy0NbEsgL.png"],"publishedAt":"2026-06-20T20:08:07.000Z","submittedOnDailyAt":"2026-07-01T00:00:00.000Z","title":"Lexical Consensus: Grounded Word Learning and Shared Meaning in Artificial Agents","submittedOnDailyBy":{"_id":"692ddcf98c9d857917c746f0","avatarUrl":"/avatars/95759d8ec90a191cf495fcbc706d244b.svg","isPro":false,"fullname":"Patricio Vera","user":"MrPatoVera","type":"user","name":"MrPatoVera"},"summary":"Artificial intelligence systems are commonly evaluated through task performance and behavioral imitation, but such evaluations leave open whether an artificial agent can acquire, stabilize, and use new lexical meanings from grounded experience. This paper introduces Lexical Consensus, an experimental framework for studying grounded word learning over a structured perceptual substrate. Using frozen DINOv2 visual embeddings, Carroll-style nonce words, and interpretable lexical learners plus linear baselines, we test whether agents can acquire artificial labels for visual concepts, generalize them bidirectionally, and stabilize them across controlled settings.\n The main result is a robust perceptual-coherence gradient: native categories are easiest to learn, coherent overextensions remain learnable, mid-range disjunctive concepts degrade, and far-disjunctive concepts approach chance. A pre-registered CIFAR-100 dissociation experiment confirms that this gradient is governed by perceptual distance rather than semantic relatedness: perceptual distance predicts acquisition accuracy (partial R^2 = 0.245, p < 1e-7), while semantic distance adds no significant explanatory power (partial R^2 = 0.002, p = 0.660).\n Bidirectional evaluation shows that naming and retrieval are distinct: exemplar-based mechanisms outperform centroid prototypes in label-to-image retrieval, exposing a memory-fidelity dimension separate from naming accuracy. Falsification controls, homogeneous candidate-pool evaluations, and null results on representational restructuring indicate that frozen perceptual geometry both enables lexical grounding and limits what can be acquired without representational adaptation.","upvotes":1,"discussionId":"6a452f104f1dd35e48fb8d20","projectPage":"https://gist.science/paper/2606.22207","githubRepo":"https://github.com/patriciomvera/lexical-consensus","githubRepoAddedBy":"user","ai_summary":"Grounded word learning experiments using visual embeddings and lexical learners reveal that perceptual distance, rather than semantic relatedness, determines acquisition success, with distinct patterns in naming and retrieval performance.","ai_keywords":["DINOv2","nonce words","lexical learners","perceptual coherence","perceptual distance","semantic distance","label-to-image retrieval","exemplar-based mechanisms","centroid prototypes","representational restructuring"],"ai_summary_model":"Qwen/Qwen2.5-Coder-32B-Instruct","githubStars":0},"canReadDatabase":false,"canManagePapers":false,"canSubmit":false,"hasHfLevelAccess":false,"upvoted":false,"upvoters":[{"_id":"692ddcf98c9d857917c746f0","avatarUrl":"/avatars/95759d8ec90a191cf495fcbc706d244b.svg","isPro":false,"fullname":"Patricio Vera","user":"MrPatoVera","type":"user"}],"acceptLanguages":["en"],"dailyPaperRank":0,"markdownContentUrl":"https://huggingface.co/buckets/huggingchat/papers-content/resolve/2606/2606.22207.md","query":{}}">
Lexical Consensus: Grounded Word Learning and Shared Meaning in Artificial Agents
Abstract
Grounded word learning experiments using visual embeddings and lexical learners reveal that perceptual distance, rather than semantic relatedness, determines acquisition success, with distinct patterns in naming and retrieval performance.
Artificial intelligence systems are commonly evaluated through task performance and behavioral imitation, but such evaluations leave open whether an artificial agent can acquire, stabilize, and use new lexical meanings from grounded experience. This paper introduces Lexical Consensus, an experimental framework for studying grounded word learning over a structured perceptual substrate. Using frozen DINOv2 visual embeddings, Carroll-style nonce words, and interpretable lexical learners plus linear baselines, we test whether agents can acquire artificial labels for visual concepts, generalize them bidirectionally, and stabilize them across controlled settings.
The main result is a robust perceptual-coherence gradient: native categories are easiest to learn, coherent overextensions remain learnable, mid-range disjunctive concepts degrade, and far-disjunctive concepts approach chance. A pre-registered CIFAR-100 dissociation experiment confirms that this gradient is governed by perceptual distance rather than semantic relatedness: perceptual distance predicts acquisition accuracy (partial R^2 = 0.245, p < 1e-7), while semantic distance adds no significant explanatory power (partial R^2 = 0.002, p = 0.660).
Bidirectional evaluation shows that naming and retrieval are distinct: exemplar-based mechanisms outperform centroid prototypes in label-to-image retrieval, exposing a memory-fidelity dimension separate from naming accuracy. Falsification controls, homogeneous candidate-pool evaluations, and null results on representational restructuring indicate that frozen perceptual geometry both enables lexical grounding and limits what can be acquired without representational adaptation.
Community
Lexical Consensus studies how artificial agents can acquire novel word--concept mappings from limited grounded visual examples. Using frozen DINOv2-small embeddings, Carroll-style artificial labels, few-shot episodes, bidirectional naming/retrieval tests, falsification controls, and multi-agent consensus experiments, the paper shows that grounded lexical acquisition is governed primarily by perceptual coherence rather than arbitrary label memorization or semantic relatedness alone. Code and experiment artifacts are available in the linked GitHub repository.
This is an automated message from the Librarian Bot. I found the following papers similar to this paper.
The following papers were recommended by the Semantic Scholar API
Please give a thumbs up to this comment if you found it helpful!
If you want recommendations for any Paper on Hugging Face checkout this Space
You can directly ask Librarian Bot for paper recommendations by tagging it in a comment: @librarian-bot recommend
Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.
Tap or paste here to upload images
Cite arxiv.org/abs/2606.22207 in a model README.md to link it from this page.
Cite arxiv.org/abs/2606.22207 in a dataset README.md to link it from this page.
Cite arxiv.org/abs/2606.22207 in a Space README.md to link it from this page.
Discussion (0)
Sign in to join the discussion. Free account, 30 seconds — email code or GitHub.
Sign in →No comments yet. Sign in and be the first to say something.