LLMs outperform single human raters at spotting relative weaknesses in L2 writing profiles on the ICNALE GRA dataset while humans are better at spotting strengths, using a self-referential intra-learner evaluation method.
hub
Advances in Neural Information Processing Systems , volume=
16 Pith papers cite this work. Polarity classification is still indexing.
16
Pith papers citing it
hub tools
citation-role summary
background 1
citation-polarity summary
roles
background 1polarities
unclear 1representative citing papers
EPIC trains LLMs to treat continuous embeddings as in-context prompts, yielding state-of-the-art text embedding performance on MTEB with or without prompts at inference and lower compute.
Time-LLM reprograms frozen LLMs for time series forecasting via text prototypes and Prompt-as-Prefix, outperforming specialized models in standard, few-shot, and zero-shot settings.
LLMs categorize rather than forecast conflict escalation across media coverage tiers, matching trivial baselines while a no-country logistic regression achieves higher F1 on a 660-case test set.
Proposes nearly balanced TCARDs that minimize the first two generalized word-length pattern components, defines Φ_BCD criterion linked to classical optimality, and constructs designs via coordinate exchange with simulation-calibrated weights for LLM prompt engineering.
SAGE is a self-evolving agentic graph-memory engine that dynamically constructs and refines structured memory graphs via writer-reader feedback, yielding performance gains on multi-hop QA, open-domain retrieval, and long-term agent benchmarks.
Recursive generative retraining with heterogeneous rewards converges to a stable distribution satisfying a weighted Nash bargaining solution, preserving diversity under stated conditions.
Response times modeled as drift-diffusion processes enable consistent estimation of population-average preferences from heterogeneous anonymous binary choices.
Adam's adaptive preconditioning and first-moment averaging improve high-probability tracking error in noise-dominated nonstationary regimes but can increase it under strong drift, where SGD achieves a smaller floor, with explicit beta-dependent bounds.
DIYHealth Suite introduces a large home-care dataset, DIYHealthGPT model with Hybrid Hyper Low-Rank Adaptation, and DIYHealthBench, claiming SOTA results on 11 tasks over general and medical baselines.
Knowledge graph triples capture 2-3x more attention per token than equivalent natural language due to structural patterns, compressing demonstration attention by up to 42% independent of semantic relevance.
A neural estimator trained on self-computed mutual information from masked diffusion model hidden states predicts the full pairwise MI matrix in one forward pass to enable faster parallel decoding of conditionally independent variables.
Empirical analysis shows scaling inference compute via strategies like tree search can be more efficient than scaling model parameters, with 7B models plus novel search outperforming 34B models.
A reference-free proxy scoring framework combined with GIRB calibration produces better-aligned evaluation metrics for summarization and outperforms baselines across seven datasets.
A reinforcement learning policy learns to adaptively harvest data samples, improving empirical constraint satisfaction and training efficiency for Lyapunov NNs and PINNs.
MiniGPT is a self-contained PyTorch implementation of standard GPT autoregressive modeling that reaches 1.478 validation loss on Tiny Shakespeare with a 10.77M-parameter model and produces recognizable Shakespeare-style text.
citing papers explorer
-
Towards Self-Referential Analytic Assessment: A Profile-Based Approach to L2 Writing Evaluation with LLMs
LLMs outperform single human raters at spotting relative weaknesses in L2 writing profiles on the ICNALE GRA dataset while humans are better at spotting strengths, using a self-referential intra-learner evaluation method.
-
Embedding-based In-Context Prompt Training for Enhancing LLMs as Text Encoders
EPIC trains LLMs to treat continuous embeddings as in-context prompts, yielding state-of-the-art text embedding performance on MTEB with or without prompts at inference and lower compute.
-
Time-LLM: Time Series Forecasting by Reprogramming Large Language Models
Time-LLM reprograms frozen LLMs for time series forecasting via text prototypes and Prompt-as-Prefix, outperforming specialized models in standard, few-shot, and zero-shot settings.
-
The Limits of LLM Forecasting: Parametric Knowledge Gaps Across Conflict Zones
LLMs categorize rather than forecast conflict escalation across media coverage tiers, matching trivial baselines while a no-country logistic regression achieves higher F1 on a 660-case test set.
-
TCARD: Nearly Balanced Two-Level Designs with Treatment Cardinality Constraints with an Application to LLM Prompt Engineering
Proposes nearly balanced TCARDs that minimize the first two generalized word-length pattern components, defines Φ_BCD criterion linked to classical optimality, and constructs designs via coordinate exchange with simulation-calibrated weights for LLM prompt engineering.
-
SAGE: A Self-Evolving Agentic Graph-Memory Engine for Structure-Aware Associative Memory
SAGE is a self-evolving agentic graph-memory engine that dynamically constructs and refines structured memory graphs via writer-reader feedback, yielding performance gains on multi-hop QA, open-domain retrieval, and long-term agent benchmarks.
-
Curated Synthetic Data Doesn't Have to Collapse: A Theoretical Study of Generative Retraining with Pluralistic Preferences
Recursive generative retraining with heterogeneous rewards converges to a stable distribution satisfying a weighted Nash bargaining solution, preserving diversity under stated conditions.
-
Response Time Enhances Alignment with Heterogeneous Preferences
Response times modeled as drift-diffusion processes enable consistent estimation of population-average preferences from heterogeneous anonymous binary choices.
-
Adapt or Forget: Provable Tradeoffs Between Adam and SGD in Nonstationary Optimization
Adam's adaptive preconditioning and first-moment averaging improve high-probability tracking error in noise-dominated nonstationary regimes but can increase it under strong drift, where SGD achieves a smaller floor, with explicit beta-dependent bounds.
-
DIYHealth Suite: Dataset, Model, and Benchmark for Health Management at Home
DIYHealth Suite introduces a large home-care dataset, DIYHealthGPT model with Hybrid Hyper Low-Rank Adaptation, and DIYHealthBench, claiming SOTA results on 11 tasks over general and medical baselines.
-
The Structural Attention Tax: How Retrieval Format Hijacks In-Context Learning Independent of Content
Knowledge graph triples capture 2-3x more attention per token than equivalent natural language due to structural patterns, compressing demonstration attention by up to 42% independent of semantic relevance.
-
Neural Estimation of Pairwise Mutual Information in Masked Discrete Sequence Models
A neural estimator trained on self-computed mutual information from masked diffusion model hidden states predicts the full pairwise MI matrix in one forward pass to enable faster parallel decoding of conditionally independent variables.
-
Inference Scaling Laws: An Empirical Analysis of Compute-Optimal Inference for Problem-Solving with Language Models
Empirical analysis shows scaling inference compute via strategies like tree search can be more efficient than scaling model parameters, with 7B models plus novel search outperforming 34B models.
-
Calibrating Model-Based Evaluation Metrics for Summarization
A reference-free proxy scoring framework combined with GIRB calibration produces better-aligned evaluation metrics for summarization and outperforms baselines across seven datasets.
-
Adaptive Data Harvesting for Efficient Neural Network Learning with Universal Constraints
A reinforcement learning policy learns to adaptively harvest data samples, improving empirical constraint satisfaction and training efficiency for Lyapunov NNs and PINNs.
-
MiniGPT: Rebuilding GPT from First Principles
MiniGPT is a self-contained PyTorch implementation of standard GPT autoregressive modeling that reaches 1.478 validation loss on Tiny Shakespeare with a 10.77M-parameter model and produces recognizable Shakespeare-style text.