Jensen-Shannon
No mentions found
This entity hasn't been tracked yet, or Iris is still building its knowledge base.
Related Articles from SNS
SHALA-LLM: Smartly Handling Ambiguous Labels in Aligning LLMs
arXiv:2606.05376v1 Announce Type: new Abstract: Many human-centered tasks, including natural language inference (NLI) and emotion recognition (ER), have multiple plausible interpretations, leading to label ambiguity and challenging disagreements across human annotators. As LLMs are increasingly deployed in real-world settings, faithfully modeling such ambiguity is essential to identify contested inputs, preserve variability in ambiguous cases, and capture the full distribution of human...
Teaching the Way, Not the Answer: Privileged Tutoring Distillation for Multimodal Policy Optimization
arXiv:2606.07000v1 Announce Type: new Abstract: Recent post-training methods, particularly Reinforcement Learning with Verifiable Rewards (RLVR), have significantly enhanced the reasoning ability of Large Vision-Language Models (LVLMs). However, the sparse nature of verifiable rewards provides little token-level supervision for failed rollouts, often leading to inefficient exploration in complex multimodal reasoning tasks.
SPECTRA: Revealing the Full Spectrum of User Preferences via Distributional LLM Inference
arXiv:2509.24189v4 Announce Type: replace Abstract: Large Language Models (LLMs) are increasingly used to model user preferences, with the typical output as a directly-generated ranked item list per user. However, this generative paradigm inherits the bias and opacity of autoregressive decoding. It over-emphasizes frequent (head) preferences and suppresses minority, long-tail ones.
RealClawBench: Live OpenClaw Benchmarks from Real Developer-Agent Sessions
arXiv:2606.03889v1 Announce Type: new Abstract: Agent benchmarks should reflect what users actually ask deployed agents to do, yet existing benchmarks often miss key realism properties of real developer-agent sessions. We introduce RealClawBench, a live benchmark framework built from real OpenClaw sessions to capture the distribution, diversity, and real-world difficulty of deployed agent use. Real user requests are challenging to benchmark because they often depend on local execution...
Reinforcement Learning from Rich Feedback with Distributional DAgger
Announce Type: new Abstract: Reasoning models have advanced rapidly, but the dominant reinforcement learning from verifiable rewards (RLVR) recipe remains surprisingly narrow: sample many responses and reward each with a single bit indicating whether the final answer is correct. Yet many settings provide rich feedback, including execution traces, tool outputs, expert corrections, and model self-evaluations. We study how to use such feedback through a distributional variant of the classic...
RealClawBench: Live OpenClaw Benchmarks from Real Developer-Agent Sessions
Announce Type: replace Abstract: Agent benchmarks should reflect what users actually ask deployed agents to do, yet existing benchmarks often miss key realism properties of real developer-agent sessions. We introduce RealClawBench, a live benchmark framework built from real OpenClaw sessions to capture the distribution, diversity, and real-world difficulty of deployed agent use. Real user requests are challenging to benchmark because they often depend on local execution environments, involve...
Reinforcement Learning from Rich Feedback with Distributional DAgger
arXiv:2606.05152v2 Announce Type: replace Abstract: Reasoning models have advanced rapidly, but the dominant reinforcement learning from verifiable rewards (RLVR) recipe remains surprisingly narrow: sample many responses and reward each with a single bit indicating whether the final answer is correct. Yet many settings provide rich feedback, including execution traces, tool outputs, expert corrections, and model self-evaluations. We study how to use such feedback through a distributional...
Can LLMs understand LilyPond? A benchmark for symbolic music generation and understanding
Announce Type: new Abstract: Symbolic music evaluation for large language models remains fragmented across representations, datasets, and metrics. We introduce LilyBench, a LilyPond-based benchmark that jointly evaluates symbolic music generation and music understanding on the same family of open-weight LLMs. The benchmark includes a 200-prompt generation suite and ten understanding tasks adapted from ABC-Eval, covering syntax, metadata prediction, structural sequencing, and music recognition.
LoTUS: Large-Scale Machine Unlearning with a Taste of Uncertainty
Announce Type: replace Abstract: We present LoTUS, a novel Machine Unlearning (MU) method that eliminates the influence of training samples from pre-trained models, avoiding retraining from scratch. LoTUS smooths the prediction probabilities of the model up to an information-theoretic bound, mitigating its over-confidence stemming from data memorization. We evaluate LoTUS on Transformer and ResNet18 models against eight baselines across five public datasets.
Beyond Individual Personas: Aligning Synthetic Dialogue to Population-Level Behavior Distributions
arXiv:2606.07893v1 Announce Type: new Abstract: Synthetic dialogue corpora are increasingly used as proxies for target dialogue data, yet persona-grounded generators optimize individual conversations rather than corpus composition, yielding locally plausible dialogues with distorted population-level behavior mixes. We introduce GroupPersona, a framework that aligns synthetic dialogue corpora to the behavior distribution of a reference corpus. GroupPersona turns population statistics into...