{"total":14,"items":[{"citing_arxiv_id":"2606.28057","ref_index":32,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"MultiHashFormer: Hash-based Generative Language Models","primary_cat":"cs.CL","submitted_at":"2026-06-26T13:03:29+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":7.0,"formal_verification":"none","one_line_summary":"MultiHashFormer enables hash-based autoregression in LMs by encoding tokens as multi-hash signatures, outperforming standard Transformers at 100M-3B scales while keeping parameter count constant for multilingual expansion.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2606.25821","ref_index":30,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"SARA: Unlocking Multilingual Knowledge in Mixture-of-Experts via Semantically Anchored Routing Alignment","primary_cat":"cs.CL","submitted_at":"2026-06-24T13:36:46+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"SARA aligns internal routing distributions in MoE layers to high-resource semantic anchors via symmetric JS divergence, improving low-resource language performance by 0.8-1.2% over standard instruction tuning on Global-MMLU.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2606.23566","ref_index":60,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"LangMAP: A Language-Adaptive Approach to Tokenization","primary_cat":"cs.CL","submitted_at":"2026-06-22T16:32:00+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":7.0,"formal_verification":"none","one_line_summary":"LangMAP adapts UnigramLM for multilingual use to deliver language-specific tokenization from a shared vocabulary, boosting boundary alignment metrics across natural and programming languages with mixed downstream fine-tuning gains.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2606.22722","ref_index":24,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"moBERTo: A Modern Encoder for Portuguese via Continued Pretraining of ModernBERT","primary_cat":"cs.CL","submitted_at":"2026-06-21T23:44:47+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":5.0,"formal_verification":"none","one_line_summary":"Continued pretraining of ModernBERT on curated Portuguese data produces moBERTo, which reports top results on Portuguese retrieval reranking and PLUE-PT benchmarks.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2606.20945","ref_index":16,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"Grouped Query Experts: Mixture-of-Experts on GQA Self-Attention","primary_cat":"cs.LG","submitted_at":"2026-06-18T21:18:42+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":7.0,"formal_verification":"none","one_line_summary":"Grouped Query Experts applies per-group MoE routing to query heads in GQA, matching baseline accuracy while activating half the query heads on 250M models trained for 30B tokens.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2606.08545","ref_index":25,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"Ishigaki-IDS: An Open-Weight Verifier-Aware Model for Information Delivery Specification Drafting in Building Information Modeling","primary_cat":"cs.CL","submitted_at":"2026-06-07T09:55:59+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":5.0,"formal_verification":"none","one_line_summary":"Ishigaki-IDS is a verifier-aware LLM for generating validator-passing IDS files in BIM, reaching IDSAuditPass scores of 0.651-0.753 on a 166-case benchmark and cutting practitioner work time by 54.7%.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.25263","ref_index":16,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"Mimir: Large-scale Multilingual Concept Modeling","primary_cat":"cs.CL","submitted_at":"2026-05-24T21:26:47+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":4.0,"formal_verification":"none","one_line_summary":"Mimir is a 1.6B multilingual concept model pretrained on 38.9 billion sentences across 46 languages and instruction-tuned on 66.8 million sentences across 35 languages, then compared to a token-based LM of similar size.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.23885","ref_index":27,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"Multilingual Knowledge Transfer under Data Constraints via Lexical Interventions","primary_cat":"cs.CL","submitted_at":"2026-05-22T17:45:49+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"LINK improves cross-lingual knowledge transfer via lexical substitutions in English pretraining data, yielding notable downstream gains and up to 2x training speedup across eight languages and five model sizes.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.18232","ref_index":22,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"SomaliWeb v1: A Quality-Filtered Somali Web Corpus with a Matched Tokenizer and a Public Language-Identification Benchmark","primary_cat":"cs.CL","submitted_at":"2026-05-18T11:28:03+00:00","verdict":"ACCEPT","verdict_confidence":"MODERATE","novelty_score":6.0,"formal_verification":"none","one_line_summary":"SomaliWeb v1 delivers a cleaned Somali corpus, efficient BPE tokenizer, and side-by-side language identification benchmark while documenting defects in prior multilingual datasets.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.18083","ref_index":24,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"A Data-Efficient Path to Multilingual LLMs: Language Expansion via Post-training PARAM$\\Delta$ Integration into Upcycled MoE","primary_cat":"cs.CL","submitted_at":"2026-05-18T08:59:08+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":7.0,"formal_verification":"none","one_line_summary":"PARAMΔ upcycles dense models to MoE for per-language experts and grafts post-training deltas to enable data-efficient language expansion while preserving original capabilities.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.13521","ref_index":14,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"Granite Embedding Multilingual R2 Models","primary_cat":"cs.IR","submitted_at":"2026-05-13T13:37:45+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":4.0,"formal_verification":"none","one_line_summary":"Granite Embedding Multilingual R2 releases 311M and 97M parameter bi-encoder models that achieve state-of-the-art retrieval performance on multilingual text, code, long-document, and reasoning datasets.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.13225","ref_index":18,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"Mix, Don't Tune: Bilingual Pre-Training Outperforms Hyperparameter Search in Data-Constrained Settings","primary_cat":"cs.LG","submitted_at":"2026-05-13T09:17:51+00:00","verdict":"CONDITIONAL","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"Mixing auxiliary high-resource language data outperforms hyperparameter tuning in data-constrained bilingual pre-training, with gains equivalent to 2-13 times more unique target data.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.12715","ref_index":34,"ref_count":2,"confidence":0.9,"is_internal_anchor":false,"paper_title":"Scaling Laws for Mixture Pretraining Under Data Constraints","primary_cat":"cs.LG","submitted_at":"2026-05-12T20:22:45+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"Empirical study shows mixture pretraining tolerates higher target data repetition than single-source training, with a new repetition-aware scaling law enabling principled mixture selection based on data size, compute, and model scale.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2508.06471","ref_index":27,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"GLM-4.5: Agentic, Reasoning, and Coding (ARC) Foundation Models","primary_cat":"cs.CL","submitted_at":"2025-08-08T17:21:06+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":4.0,"formal_verification":"none","one_line_summary":"GLM-4.5, a 355B-parameter MoE model with hybrid reasoning, scores 70.1% on TAU-Bench, 91.0% on AIME 24, and 64.2% on SWE-bench Verified while ranking 3rd overall and 2nd on agentic benchmarks.","context_count":1,"top_context_role":"dataset","top_context_polarity":"use_dataset","context_text":"Such webpages cannot be removed by MinHash deduplication. We additionally apply the SemDedup [1] pipeline to remove those similar webpages based on document embeddings. Multilingual To support more natural languages, we include multilingual documents in our pre- training corpus. The multilingual corpus comes from both our crawled webpages and Fineweb-2 [27]. We apply a quality classifier that judges the educational utility of documents and up-sample high- quality multilingual documents. Code We curated source code data from GitHub and various code hosting platforms. The code corpus undergoes a preliminary rule-based filtering, followed by classification using language-specific quality models that categorize samples into three tiers: high-quality, medium-quality, and low-quality."}],"limit":50,"offset":0}