Author

D. Bikel

h-index 19 View on Semantic Scholar

We have 3 of 60 papers

We haven’t gathered this author’s papers yet. Follow them and we’ll fetch their work.

Not the right person? Other researchers publish under this name.

#artificial intelligence Preprint Oct 2026

Denoising Blocks, Not Tokens: Efficient Compressed Continuous Diffusion with Branching Token Realization

Diffusion language models (DLMs) generate text through iterative parallel refinement, offering the potential for higher throughput than autoregressive (AR) decoding. However, most DLMs still maintain one generative state per token, so every denoising step processes a state sequence as long as the output sequence, limit...

Xin-Song Feng, Peng Du, Zhi-Zhuo Yang et al. · 0 citations

#artificial intelligence Preprint Aug 2026

Palmyra x6 Technical Report: An Agentic, Tool-Use Model Post-Trained via Anchored Supervised Fine-Tuning

The model shows substantial gains over the previous default model for Writer Agent, and compares favorably with several recent models on public benchmarks, scoring the highest on BFCL Core at $0.785$ and posts the highest six-benchmark mean of the cohort.

Peng Du, Kiran P. Kamble, R. Vasudev et al. · 0 citations

#natural language process... May 2026

Auditing LLM Benchmarks with Item Response Theory

An Item Response Theory-based indicator is introduced that surfaces likely mislabels at 95% precision in the top 200 examples across seven preference and multiple-choice benchmarks using responses from 114 models, outperforming a supervised classifier.

Sander Land, D. Bikel · arXiv.org · 2 citations