Claims Browser

AI watermarking is most likely implemented by biasing random token selection to create a detectable pattern in token IDs

"the most likely implementation is that they bias the random token selection such that there's a detectable pattern in the token IDs"

TabuLM is the first language model pre-trained on Kinyarwanda tabular data

"We present TabuLM, the first language model pre-trained on Kinyarwanda tabular data."

TabuLM achieves 62.0% exact match on TabQA-kin, outperforming KinyaBERT-large by 5.7 EM points and multilingual baselines by 11.7-12.7 points

"TabuLM achieves 62.0% exact match on TabQA-kin, outperforming KinyaBERT-large by 5.7 EM points and all multilingual baselines (mBERT 49.3%, XLM-R 50.0%) by 11.7-12.7 points."

Structural table embeddings are most decisive for comparison and lookup questions, while morphological awareness provides complementary gains

"Analysis shows that structural table embeddings are most decisive for comparison and lookup questions, while morphological awareness provides complementary gains."

Cross-lingual word-to-speech mappings can be learned directly from visual grounding without transcriptions or explicit model training

"Our work demonstrates that cross-lingual word-to-speech mappings can be learned directly from visual grounding without transcriptions or explicit model training."

Alignment-based approach using self-supervised features outperforms attention-based neural models for keyword spotting and localization

"Experiments evaluating keyword spotting and localization show that our alignment-based approach outperforms a previous attention-based neural model."

KinyaEmbed is the first dedicated sentence embedding model for Kinyarwanda

"We present KinyaEmbed, the first dedicated sentence embedding model for Kinyarwanda, a morphologically rich Bantu language spoken by over 12 million people in Rwanda."

Existing multilingual embedding models perform poorly on Kinyarwanda due to severe under-representation in their pre-training corpora

"Existing multilingual embedding models such as LaBSE, mE5-large, and OpenAI text-embedding-3-large perform poorly on Kinyarwanda due to severe under-representation in their pre-training corpora."

KinyaEmbed achieves Spearman correlation of 0.7298 on SemRel2024-rw, surpassing mE5-large by 20.9% and OpenAI text-embedding-3-large by 41.0%

"A seven-checkpoint ensemble (all5+23A*2, with the final stage double-weighted) achieves Spearman \r{ho}=0.7298 on SemRel2024-rw, surpassing mE5-large by 20.9% and OpenAI text-embedding-3-large by 41.0%."

Combining procedural generation with 4D visual dynamics in one environment still demands extensive manual effort and results are rarely editable or controllable enough to reuse at scale

"Combining these properties in one environment, however, still demands extensive manual effort, and the result is rarely editable or controllable enough to reuse at scale."

Vision-language models fail the majority of navigation tasks in 4DSynth-Nav benchmark and stall after early subtasks

"Two vision-language models evaluated across three difficulty tiers both fail the majority of tasks and stall after early subtasks."

Current 3DGS compression systems combine multiple strategies which can obscure where gains come from and limit component reuse across training pipelines

"Current 3DGS compression systems combine multiple strategies for file size reduction, which can obscure where gains come from and limit component reuse across training pipelines."

KISS-GS achieves 85x to 319x reductions in scene size over uncompressed vanilla 3DGS, setting new benchmarks for real-world scenes

"Across standard 3DGS benchmarks, our simple and modular approach thus achieves a total of 85x to 319x reductions in the size of the scene over uncompressed vanilla 3DGS, setting new benchmarks for real-world scenes and surpassing tightly integrated methods in rate-distortion."

Most existing math benchmarks evaluate only final answers, providing limited diagnostic value for identifying process-level failures

"However, most existing math benchmarks evaluate only final answers. This outcome-oriented evaluation provides limited diagnostic value for identifying process-level failures or rigorous logic, failing to guide the transformation of LLMs into robust agents."

Models with similar end-to-end accuracy can exhibit markedly different agentic capability profiles

"Experiments reveal that models with similar end-to-end accuracy can exhibit markedly different agentic capability profiles."

Process-level evaluation is crucial for interpreting the true potential of LLMs and guiding development of next-generation mathematical agents

"This demonstrates that process-level evaluation is crucial for interpreting the true potential of LLMs and guiding the development of next-generation mathematical agents."

Casual captures frequently contain transient objects that can produce blurred, duplicated, or floating artifacts in 3D Gaussian Splatting novel views

"However, such captures frequently contain transient objects that appear in only a subset of the views. Such content can be encoded into the per-view Gaussians associated with the inputs that observe it and remain in the combined representation despite being observed by no other input. As a result, it may produce blurred, duplicated, or floating artifacts in novel views."

Most existing visual reward models use a single scalar score or rely on fixed criteria that cannot adapt to different instructions

"Reward models play an essential role in aligning visual generative models, yet most existing visual reward models use a single scalar score or rely on fixed criteria that cannot adapt to different instructions. This limits both interpretability and task sensitivity, especially for text-to-image generation and instruction-based image editing, where different inputs require different evaluation dimensions."