Human Feedback and Eval Paper Explorer

Match reason: Ranked by recency.

• Our contributions are three-fold: (1) a structured representation framework for situated lexical meaning; (2) COCA-Scenes, a dataset of 520 usage instances across 26 keywords for distinct scene identification; and (3) empirical evidence…

Match reason: Ranked by recency.

• The annotation process combines expert human judgment with model-assisted pre-labeling verified by trained annotators, achieving substantial inter-annotator agreement (Cohens kappa = 0.85).

Match reason: Ranked by recency.

• Large language model (LLM) agents still struggle with long-term memory question answering, where answer-supporting evidence is often scattered across long conversational histories and buried in substantial irrelevant content.

Match reason: Ranked by recency.

• Multimodal Large Language Models (MLLMs) have made rapid progress in spatial intelligence, yet existing spatial reasoning benchmarks largely assume pristine visual inputs and overlook the degradations that commonly occur in real-world…
• Finally, we show that finetuning on SpaceDG markedly improves degradation robustness and can even surpass human performance under degraded conditions without any performance drop on clean images, highlighting the promise of…

Match reason: Ranked by recency.

• Our method demonstrates strong performance on multiple BEL benchmarks, yielding significant improvements in linking accuracy (3%-24%) while reducing inference time compared to the state-of-the-art.

Match reason: Ranked by recency.

• Recent task-specific attention operators can compress deep Transformer stacks into a few layers by performing multi-hop state propagation within a single layer, but their dense evaluation remains expensive.
• On controlled tracking benchmarks, our method matches the dense operator's accuracy while reducing wall-clock time by 12-29\% under a standardized measurement protocol, and is up to 2.4 \times faster than a compact dense Transformer at…

Match reason: Ranked by recency.

• Two NLA-inspired evaluations strengthen this picture: the fifteen selective features explain only 31% of activation variance versus the SAE's 99.7%, and selectivity ratio anticorrelates with causal force (r = -0.56).
• A cost-based deployment evaluation (assumed 50/FN, 0.42/FP, 2% error rate) finds an optimal monitor configuration yielding 8.96 per 1000 queries against a 1000 baseline, a 99.1% saving.

Match reason: Ranked by recency.

• Empirical results on challenging multilingual mathematical benchmarks reveal that LANG substantially enhances reasoning performance without compromising language consistency.

Match reason: Ranked by recency.

• Post-training has become the dominant recipe for turning a language model into a competent search-augmented reasoning agent.
• We take a step back and ask whether any of this machinery is actually necessary, and propose Search-E1, a self-evolution method that lets a search-augmented agent improve through only vanilla GRPO interleaved with offline self-distillation…

Match reason: Ranked by recency.

• Abstract shows limited direct human-feedback or evaluation-protocol detail; use as adjacent methodological context.

Match reason: Ranked by recency.

• Abstract shows limited direct human-feedback or evaluation-protocol detail; use as adjacent methodological context.

Match reason: Ranked by recency.

• Abstract shows limited direct human-feedback or evaluation-protocol detail; use as adjacent methodological context.

Match reason: Ranked by recency.

• We introduce SynAE, an evaluation framework for assessing how well synthetic benchmarks for multi-turn, tool-calling agents replicate and augment the characteristics of real data trajectories.
• We evaluate SynAE using recent agent benchmarks and test common synthetic data failure modes via realistic and controlled generation schemes.

Match reason: Ranked by recency.

• The single-point evaluation ignores a main problem of the instruction-based approach namely: sensitivity to the phrasing of the instruction.
• Our findings suggest that single-prompt evaluation is insufficient for instruction-tuned embedding models and that benchmarks should incorporate prompt robustness, either by evaluating over multiple prompts or by reporting sensitivity…

Match reason: Ranked by recency.

• In this study, we investigate an agent designed to encourage integration by adapting to the individual user's thought patterns.
• We explore its effects on participants' perceptions of the agent and their reflective behavior, in comparison with unaided pre-reflection and a baseline agent.

Match reason: Ranked by recency.

• To address this limitation, we introduce a novel, culturally aligned instruction-tuning dataset for BangLa Application and DialoguE generation - BLADE and benchmarking framework comprising 4,196 meticulously curated interaction pairs.
• Our empirical evaluations demonstrate that models fine-tuned on our dataset yield substantial improvements in structural fidelity and honorific alignment, providing a rigorous benchmark for bridging pragmatic disparities in low-resource…

Match reason: Ranked by recency.

• 23 experts revise the generated CS, which are assessed via human and automatic metrics.
• Based on the post-edited CS, the mixed strategy proves to be the most effective in crowdsourcing evaluation, pairing strong factual correction with stereotype mitigation and empathetic engagement.

Match reason: Ranked by recency.

• Abstract shows limited direct human-feedback or evaluation-protocol detail; use as adjacent methodological context.

Match reason: Ranked by recency.

• Abstract shows limited direct human-feedback or evaluation-protocol detail; use as adjacent methodological context.

Match reason: Ranked by recency.

• Abstract shows limited direct human-feedback or evaluation-protocol detail; use as adjacent methodological context.