LLMs fail causal discovery due to a kernel obstruction in observational learning, but interventional agents using frozen LLMs in Bayesian loops succeed without training on causal graph benchmarks.
Title resolution pending
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
verdicts
UNVERDICTED 3representative citing papers
BLINC uses large language models to guide Bayesian network causal learning for RAN parameter optimization, delivering 63.5% throughput gains and 19.7% block error rate reduction over data-only baselines in a private 5G testbed while enabling interpretable, adaptive models.
Introduces CounterBench benchmark and CoIn iterative reasoning method showing LLMs perform near random on formal counterfactual tasks but improve substantially with guided backtracking.
citing papers explorer
-
Why LLMs Fail at Causal Discovery and How Interventional Agents Escape
LLMs fail causal discovery due to a kernel obstruction in observational learning, but interventional agents using frozen LLMs in Bayesian loops succeed without training on causal graph benchmarks.
-
BLINC: Context-Specific Causal Learning for Automated RAN Configuration
BLINC uses large language models to guide Bayesian network causal learning for RAN parameter optimization, delivering 63.5% throughput gains and 19.7% block error rate reduction over data-only baselines in a private 5G testbed while enabling interpretable, adaptive models.
-
CounterBench: Evaluating and Improving Counterfactual Reasoning in Large Language Models
Introduces CounterBench benchmark and CoIn iterative reasoning method showing LLMs perform near random on formal counterfactual tasks but improve substantially with guided backtracking.