Geometry-aware FGW prior transfer plus PAC-Bayes residual adaptation reduces topology forgetting and raises average accuracy across continual multi-agent topology learning streams.
Measuring massive multitask language understanding
6 Pith papers cite this work. Polarity classification is still indexing.
6
Pith papers citing it
representative citing papers
WMF-AM is a depth-parameterized benchmark that measures LLMs' cumulative state tracking ability without scratchpads, validated on 28 models across arithmetic and non-arithmetic tasks with ablations confirming the construct.
DASD improves math reasoning in LLMs by adaptively directing self-distillation based on per-token entropy to balance exploration and step accuracy, outperforming prior self-distillation and RLVR baselines on six benchmarks.
Higher worker capability in multi-agent LLM systems increases semantic hijacking attack success rates via linguistic certainty in reports, with heterogeneous ensembles reducing ASR from 52.8% to 2.0%.
Systematic study of inter-agent communication in LLM multi-agent systems shows reasoning and verification are critical for performance, with a new augmentation technique recovering 86.2% of failures.
A 1.4B and a 2.6B looped (weight-tied, recurrent-depth) language model trained on 7.7T tokens match or exceed several 4B–8B transformer baselines on selected reasoning benchmarks.
citing papers explorer
-
MasFACT: Continual Multi-Agent Topology Learning via Geometry-Aware Posterior Transfer
Geometry-aware FGW prior transfer plus PAC-Bayes residual adaptation reduces topology forgetting and raises average accuracy across continual multi-agent topology learning streams.
-
WMF-AM: Probing LLM Working Memory via Depth-Parameterized Cumulative State Tracking
WMF-AM is a depth-parameterized benchmark that measures LLMs' cumulative state tracking ability without scratchpads, validated on 28 models across arithmetic and non-arithmetic tasks with ablations confirming the construct.
-
Tailoring Teaching to Aptitude: Direction-Adaptive Self-Distillation for LLM Reasoning
DASD improves math reasoning in LLMs by adaptively directing self-distillation based on per-token entropy to balance exploration and step accuracy, outperforming prior self-distillation and RLVR baselines on six benchmarks.
-
The Capability Paradox: How Smarter Auditors Make Multi-Agent Systems Less Secure
Higher worker capability in multi-agent LLM systems increases semantic hijacking attack success rates via linguistic certainty in reports, with heterogeneous ensembles reducing ASR from 52.8% to 2.0%.
-
What Do Agents Communicate? Characterizing Information Exchange in Multi-Agent Systems
Systematic study of inter-agent communication in LLM multi-agent systems shows reasoning and verification are critical for performance, with a new augmentation technique recovering 86.2% of failures.
-
Scaling Latent Reasoning via Looped Language Models
A 1.4B and a 2.6B looped (weight-tied, recurrent-depth) language model trained on 7.7T tokens match or exceed several 4B–8B transformer baselines on selected reasoning benchmarks.