OracleProto is a reproducible framework that uses model-cutoff alignment, temporal masking, and leakage detection to create low-leakage benchmarks for LLM native forecasting from past events.
Time-r1: Towards comprehensive temporal reasoning in llms
5 Pith papers cite this work. Polarity classification is still indexing.
5
Pith papers citing it
representative citing papers
Language models frequently violate temporal scope stability in multi-turn dialogues by drifting toward present-day assumptions even when they possess the correct facts.
An agentic forecaster with linguistic belief states, logit-space multi-trial shrinkage, and hierarchical Platt calibration achieves SOTA Brier Index on ForecastBench binary questions.
QDET deploys a 7B-parameter model fine-tuned with three auxiliary tasks and RL that matches a 671B model's F1 on query-driven timeline summarization while delivering measurable gains in production search metrics.
Self-aligned reward uses relative perplexity differences to encourage concise, query-specific reasoning in LLMs, yielding 4% accuracy gains and 30% lower inference cost when added to PPO or GRPO.
citing papers explorer
-
OracleProto: A Reproducible Framework for Benchmarking LLM Native Forecasting via Knowledge Cutoff and Temporal Masking
OracleProto is a reproducible framework that uses model-cutoff alignment, temporal masking, and leakage detection to create low-leakage benchmarks for LLM native forecasting from past events.
-
Evaluating Temporal Consistency in Multi-Turn Language Models
Language models frequently violate temporal scope stability in multi-turn dialogues by drifting toward present-day assumptions even when they possess the correct facts.
-
Agentic Forecasting using Sequential Bayesian Updating of Linguistic Beliefs
An agentic forecaster with linguistic belief states, logit-space multi-trial shrinkage, and hierarchical Platt calibration achieves SOTA Brier Index on ForecastBench binary questions.
-
Large Language Model-Powered Query-Driven Event Timeline Summarization in Industrial Search
QDET deploys a 7B-parameter model fine-tuned with three auxiliary tasks and RL that matches a 671B model's F1 on query-driven timeline summarization while delivering measurable gains in production search metrics.
-
Self-Aligned Reward: Towards Effective and Efficient Reasoners
Self-aligned reward uses relative perplexity differences to encourage concise, query-specific reasoning in LLMs, yielding 4% accuracy gains and 30% lower inference cost when added to PPO or GRPO.