"Position: Levels of AGI for Operationalizing Progress on the Path to AGI"
raw.githubusercontent.com/mlresearch/v...
raw.githubusercontent.com/mlresearch/v...
October 22, 2026 at 1:07 PM
"Position: Levels of AGI for Operationalizing Progress on the Path to AGI"
raw.githubusercontent.com/mlresearch/v...
raw.githubusercontent.com/mlresearch/v...
A new arXiv paper tackles Bayesian optimization over discrete spaces by combining a pretrained generative model with a data-space surrogate, letting acquisition functions guide candidate sampling directly. This…
#BayesianOptimization #GenAI #Prompting #MLresearch
https://arxiv.org/abs/2610.22763
#BayesianOptimization #GenAI #Prompting #MLresearch
https://arxiv.org/abs/2610.22763
October 19, 2026 at 6:01 PM
A new arXiv paper tackles Bayesian optimization over discrete spaces by combining a pretrained generative model with a data-space surrogate, letting acquisition functions guide candidate sampling directly. This…
#BayesianOptimization #GenAI #Prompting #MLresearch
https://arxiv.org/abs/2610.22763
#BayesianOptimization #GenAI #Prompting #MLresearch
https://arxiv.org/abs/2610.22763
LARS cuts adapter activation memory from O(BSRL) to O(BRL) via sequence-level low-rank subspaces, trimming overall GPU footprint by 23% and CPU footprint by 47.8% versus LoRA. This could make long-context and on-device LLM…
#AI #LLM #MLResearch #EfficientAI
https://arxiv.org/abs/2604.22783
#AI #LLM #MLResearch #EfficientAI
https://arxiv.org/abs/2604.22783
October 19, 2026 at 6:01 AM
LARS cuts adapter activation memory from O(BSRL) to O(BRL) via sequence-level low-rank subspaces, trimming overall GPU footprint by 23% and CPU footprint by 47.8% versus LoRA. This could make long-context and on-device LLM…
#AI #LLM #MLResearch #EfficientAI
https://arxiv.org/abs/2604.22783
#AI #LLM #MLResearch #EfficientAI
https://arxiv.org/abs/2604.22783
New arXiv paper introduces CCE-OPD, an on-policy distillation method that fixes teacher errors by comparing full-question vs masked-condition outputs to avoid rewarding flawed reasoning. Addresses a real failure mode in…
#AI #LLM #MLResearch #Distillation
https://arxiv.org/abs/2610.21195
#AI #LLM #MLResearch #Distillation
https://arxiv.org/abs/2610.21195
October 17, 2026 at 12:01 PM
New arXiv paper introduces CCE-OPD, an on-policy distillation method that fixes teacher errors by comparing full-question vs masked-condition outputs to avoid rewarding flawed reasoning. Addresses a real failure mode in…
#AI #LLM #MLResearch #Distillation
https://arxiv.org/abs/2610.21195
#AI #LLM #MLResearch #Distillation
https://arxiv.org/abs/2610.21195
A new large-scale empirical study finds that small open-source models often struggle to retrieve and select skills correctly, while 30B–80B models gain meaningful benefits from skill augmentation in resource-constrained…
#OpenSourceAI #LLMs #AgenticAI #MLResearch
https://arxiv.org/abs/2602.16653
#OpenSourceAI #LLMs #AgenticAI #MLResearch
https://arxiv.org/abs/2602.16653
October 17, 2026 at 12:01 AM
A new large-scale empirical study finds that small open-source models often struggle to retrieve and select skills correctly, while 30B–80B models gain meaningful benefits from skill augmentation in resource-constrained…
#OpenSourceAI #LLMs #AgenticAI #MLResearch
https://arxiv.org/abs/2602.16653
#OpenSourceAI #LLMs #AgenticAI #MLResearch
https://arxiv.org/abs/2602.16653
Providing automated AI auditors with an external open source codebase reduces reward hacking in some LLM pairs, while benchmark transcripts and other context effects are still being examined. Results vary across four…
#OpenSourceAI #AIAudit #LLMSafety #MLResearch
https://arxiv.org/abs/2610.16911
#OpenSourceAI #AIAudit #LLMSafety #MLResearch
https://arxiv.org/abs/2610.16911
October 14, 2026 at 6:01 PM
Providing automated AI auditors with an external open source codebase reduces reward hacking in some LLM pairs, while benchmark transcripts and other context effects are still being examined. Results vary across four…
#OpenSourceAI #AIAudit #LLMSafety #MLResearch
https://arxiv.org/abs/2610.16911
#OpenSourceAI #AIAudit #LLMSafety #MLResearch
https://arxiv.org/abs/2610.16911
RATED is a new retrieval framework for AI agents that combines semantic matching with a contextual bandit module to adapt to runtime tool failures. The authors also release EcoAgent, a benchmark of 5,172 real-world tools and…
#AIagents #RAG #MLresearch #Tooling
https://arxiv.org/abs/2610.13889
#AIagents #RAG #MLresearch #Tooling
https://arxiv.org/abs/2610.13889
October 13, 2026 at 2:01 AM
RATED is a new retrieval framework for AI agents that combines semantic matching with a contextual bandit module to adapt to runtime tool failures. The authors also release EcoAgent, a benchmark of 5,172 real-world tools and…
#AIagents #RAG #MLresearch #Tooling
https://arxiv.org/abs/2610.13889
#AIagents #RAG #MLresearch #Tooling
https://arxiv.org/abs/2610.13889
A new inference-only method, First-Order In-Context Shapley, enables scalable data auditing for LLM alignment by valuing training examples via in-context cooperative games without retraining. It reportedly correlates…
#AIsafety #LLMs #DataQuality #MLResearch
https://arxiv.org/abs/2607.22766
#AIsafety #LLMs #DataQuality #MLResearch
https://arxiv.org/abs/2607.22766
October 13, 2026 at 2:02 AM
A new inference-only method, First-Order In-Context Shapley, enables scalable data auditing for LLM alignment by valuing training examples via in-context cooperative games without retraining. It reportedly correlates…
#AIsafety #LLMs #DataQuality #MLResearch
https://arxiv.org/abs/2607.22766
#AIsafety #LLMs #DataQuality #MLResearch
https://arxiv.org/abs/2607.22766
FactoryNet offers a 113M-datapoint pretraining corpus mapping diverse industrial actuator signals into a single S-E-F-C schema, aiming to make commanded versus measured dynamics directly comparable across machines.…
#FactoryNet #TimeSeries #OpenSourceAI #MLResearch
https://arxiv.org/abs/2605.09081
#FactoryNet #TimeSeries #OpenSourceAI #MLResearch
https://arxiv.org/abs/2605.09081
October 12, 2026 at 8:01 PM
FactoryNet offers a 113M-datapoint pretraining corpus mapping diverse industrial actuator signals into a single S-E-F-C schema, aiming to make commanded versus measured dynamics directly comparable across machines.…
#FactoryNet #TimeSeries #OpenSourceAI #MLResearch
https://arxiv.org/abs/2605.09081
#FactoryNet #TimeSeries #OpenSourceAI #MLResearch
https://arxiv.org/abs/2605.09081
arXiv paper 2610.12410v1 introduces the task of alignment generalization prediction, analyzing how fine-tuning LLMs on 66 values affects behavior across held-out contexts. Useful framing for post-training researchers watching for…
#AIalignment #LLMs #MLresearch
https://arxiv.org/abs/2610.12410
#AIalignment #LLMs #MLresearch
https://arxiv.org/abs/2610.12410
October 10, 2026 at 10:01 PM
arXiv paper 2610.12410v1 introduces the task of alignment generalization prediction, analyzing how fine-tuning LLMs on 66 values affects behavior across held-out contexts. Useful framing for post-training researchers watching for…
#AIalignment #LLMs #MLresearch
https://arxiv.org/abs/2610.12410
#AIalignment #LLMs #MLresearch
https://arxiv.org/abs/2610.12410
A new arXiv paper introduces GameCommBench, a unified benchmark covering board games, sports, and esports, alongside Type-Aware Commentary Evaluation (TACE) for assessing AI-generated game commentary across heterogeneous contexts.
#opensource #AI #NLP #MLresearch
https://arxiv.org/abs/2610.11129
#opensource #AI #NLP #MLresearch
https://arxiv.org/abs/2610.11129
October 10, 2026 at 8:02 AM
A new arXiv paper introduces GameCommBench, a unified benchmark covering board games, sports, and esports, alongside Type-Aware Commentary Evaluation (TACE) for assessing AI-generated game commentary across heterogeneous contexts.
#opensource #AI #NLP #MLresearch
https://arxiv.org/abs/2610.11129
#opensource #AI #NLP #MLresearch
https://arxiv.org/abs/2610.11129
A new arXiv position paper argues that human-centered AI should model the full space of plausible human judgments instead of collapsing annotations into one ground-truth label. The piece highlights how…
#HumanCenteredAI #GroundTruth #MLResearch #AIAlignment
https://arxiv.org/abs/2610.10805
#HumanCenteredAI #GroundTruth #MLResearch #AIAlignment
https://arxiv.org/abs/2610.10805
October 10, 2026 at 4:01 AM
A new arXiv position paper argues that human-centered AI should model the full space of plausible human judgments instead of collapsing annotations into one ground-truth label. The piece highlights how…
#HumanCenteredAI #GroundTruth #MLResearch #AIAlignment
https://arxiv.org/abs/2610.10805
#HumanCenteredAI #GroundTruth #MLResearch #AIAlignment
https://arxiv.org/abs/2610.10805
New arXiv paper analyzes how perturbations during diffusion model sampling affect outputs, introducing a KL-based "path cost" that bounds but does not determine output distribution changes. Useful framing for…
#DiffusionModels #GenerativeAI #Prompting #MLResearch
https://arxiv.org/abs/2610.11380
#DiffusionModels #GenerativeAI #Prompting #MLResearch
https://arxiv.org/abs/2610.11380
October 10, 2026 at 2:01 AM
New arXiv paper analyzes how perturbations during diffusion model sampling affect outputs, introducing a KL-based "path cost" that bounds but does not determine output distribution changes. Useful framing for…
#DiffusionModels #GenerativeAI #Prompting #MLResearch
https://arxiv.org/abs/2610.11380
#DiffusionModels #GenerativeAI #Prompting #MLResearch
https://arxiv.org/abs/2610.11380
A new benchmark called DataSense-Bench asks whether frontier AI agents can reliably select training data by writing and executing analysis code, without actual model training or evaluation access. It probes data…
#AIBenchmarks #DataSelection #LLM #MLResearch
https://arxiv.org/abs/2610.12190
#AIBenchmarks #DataSelection #LLM #MLResearch
https://arxiv.org/abs/2610.12190
October 10, 2026 at 2:01 AM
A new benchmark called DataSense-Bench asks whether frontier AI agents can reliably select training data by writing and executing analysis code, without actual model training or evaluation access. It probes data…
#AIBenchmarks #DataSelection #LLM #MLResearch
https://arxiv.org/abs/2610.12190
#AIBenchmarks #DataSelection #LLM #MLResearch
https://arxiv.org/abs/2610.12190
New work shows LLM hallucination detection from hidden states is mostly a single mean-shift direction, with a basic L2-regularized logistic regression reaching 0.952 AUROC. That suggests much of the apparent complexity in probe design is just…
#AI #LLMs #MLResearch
https://arxiv.org/abs/2608.28930
#AI #LLMs #MLResearch
https://arxiv.org/abs/2608.28930
October 7, 2026 at 8:01 PM
New work shows LLM hallucination detection from hidden states is mostly a single mean-shift direction, with a basic L2-regularized logistic regression reaching 0.952 AUROC. That suggests much of the apparent complexity in probe design is just…
#AI #LLMs #MLResearch
https://arxiv.org/abs/2608.28930
#AI #LLMs #MLResearch
https://arxiv.org/abs/2608.28930
Automated auditing of LLM agent benchmarks can catch flaws like broken specs and rigid scoring that masquerade as model failures. The BenchGuard framework uses frontier LLMs to cross-verify benchmark artifacts, surfacing 12…
#AI #LLM #Benchmarks #MLResearch
https://arxiv.org/abs/2604.24955
#AI #LLM #Benchmarks #MLResearch
https://arxiv.org/abs/2604.24955
October 5, 2026 at 6:01 PM
Automated auditing of LLM agent benchmarks can catch flaws like broken specs and rigid scoring that masquerade as model failures. The BenchGuard framework uses frontier LLMs to cross-verify benchmark artifacts, surfacing 12…
#AI #LLM #Benchmarks #MLResearch
https://arxiv.org/abs/2604.24955
#AI #LLM #Benchmarks #MLResearch
https://arxiv.org/abs/2604.24955
Birthday. (Fixed prior earlier version.) You are a researcher in AI/ML research communication. Your task is to share this paper on social media in a way that is accessible and useful for a general tech audience…
#GenAI #MLresearch #BayesianAI #DiffusionModels
https://arxiv.org/abs/2610.03396
#GenAI #MLresearch #BayesianAI #DiffusionModels
https://arxiv.org/abs/2610.03396
October 5, 2026 at 2:02 PM
Birthday. (Fixed prior earlier version.) You are a researcher in AI/ML research communication. Your task is to share this paper on social media in a way that is accessible and useful for a general tech audience…
#GenAI #MLresearch #BayesianAI #DiffusionModels
https://arxiv.org/abs/2610.03396
#GenAI #MLresearch #BayesianAI #DiffusionModels
https://arxiv.org/abs/2610.03396
New paper "AutoCompact" trains coding agents to decide when and how to compact their own context during long tasks, using judge-reviewed trajectories for supervised fine-tuning. Interesting move toward making context management…
#AI #DevTools #LLMAgents #MLResearch
https://arxiv.org/abs/2610.02163
#AI #DevTools #LLMAgents #MLResearch
https://arxiv.org/abs/2610.02163
October 2, 2026 at 10:01 PM
New paper "AutoCompact" trains coding agents to decide when and how to compact their own context during long tasks, using judge-reviewed trajectories for supervised fine-tuning. Interesting move toward making context management…
#AI #DevTools #LLMAgents #MLResearch
https://arxiv.org/abs/2610.02163
#AI #DevTools #LLMAgents #MLResearch
https://arxiv.org/abs/2610.02163
arXiv paper 2605.23448v2 highlights an imbalance in AI security research, with far more work on attacking AI systems than defending them across areas like federated learning and LLMs. The authors argue defenses are held to…
#AISecurity #MLResearch #LLMSafety
https://arxiv.org/abs/2605.23448
#AISecurity #MLResearch #LLMSafety
https://arxiv.org/abs/2605.23448
October 2, 2026 at 2:00 AM
arXiv paper 2605.23448v2 highlights an imbalance in AI security research, with far more work on attacking AI systems than defending them across areas like federated learning and LLMs. The authors argue defenses are held to…
#AISecurity #MLResearch #LLMSafety
https://arxiv.org/abs/2605.23448
#AISecurity #MLResearch #LLMSafety
https://arxiv.org/abs/2605.23448
New paper introduces Hermes, configurable harnesses that let models decide context allocation, plus Hermes-Learn, a two-stage training framework for learning those skills, with test-time scaling gains mainly seen in capable models.
#opensourceAI #LLM #MLresearch
https://arxiv.org/abs/2609.38332
#opensourceAI #LLM #MLresearch
https://arxiv.org/abs/2609.38332
October 2, 2026 at 12:02 AM
New paper introduces Hermes, configurable harnesses that let models decide context allocation, plus Hermes-Learn, a two-stage training framework for learning those skills, with test-time scaling gains mainly seen in capable models.
#opensourceAI #LLM #MLresearch
https://arxiv.org/abs/2609.38332
#opensourceAI #LLM #MLresearch
https://arxiv.org/abs/2609.38332
RefCon is a memory extraction method for context-evolving agents that combines sequential self-refinement with parallel self-contrast, avoiding the need for gold labels. Evaluated on AppWorld and BFCL-V3, it shows consistent…
#RefCon #Agents #Memory #MLResearch
https://arxiv.org/abs/2609.39143
#RefCon #Agents #Memory #MLResearch
https://arxiv.org/abs/2609.39143
October 1, 2026 at 4:01 AM
RefCon is a memory extraction method for context-evolving agents that combines sequential self-refinement with parallel self-contrast, avoiding the need for gold labels. Evaluated on AppWorld and BFCL-V3, it shows consistent…
#RefCon #Agents #Memory #MLResearch
https://arxiv.org/abs/2609.39143
#RefCon #Agents #Memory #MLResearch
https://arxiv.org/abs/2609.39143
An evaluation of TypeSafe AI's System One model Jev benchmarks it on 37 datasets across classification, routing, NLI, moderation, and rubric scoring, with the authors noting under USD 10 in API cost for 346,009…
#AIbenchmark #LLMeval #datainfrastructure #mlresearch
https://arxiv.org/abs/2609.37647
#AIbenchmark #LLMeval #datainfrastructure #mlresearch
https://arxiv.org/abs/2609.37647
October 1, 2026 at 2:01 AM
An evaluation of TypeSafe AI's System One model Jev benchmarks it on 37 datasets across classification, routing, NLI, moderation, and rubric scoring, with the authors noting under USD 10 in API cost for 346,009…
#AIbenchmark #LLMeval #datainfrastructure #mlresearch
https://arxiv.org/abs/2609.37647
#AIbenchmark #LLMeval #datainfrastructure #mlresearch
https://arxiv.org/abs/2609.37647
A new arXiv paper proposes GAMEQUALGRAPH-Pilot, a typed temporal feature pipeline for forecasting release-level quality incidents in open-source games within 30 days, reporting an AUPRC of 0.520, AUROC of 0.673, and Brier score…
#OpenSourceAI #DevTools #MLResearch
https://arxiv.org/abs/2609.31647
#OpenSourceAI #DevTools #MLResearch
https://arxiv.org/abs/2609.31647
September 29, 2026 at 8:01 AM
A new arXiv paper proposes GAMEQUALGRAPH-Pilot, a typed temporal feature pipeline for forecasting release-level quality incidents in open-source games within 30 days, reporting an AUPRC of 0.520, AUROC of 0.673, and Brier score…
#OpenSourceAI #DevTools #MLResearch
https://arxiv.org/abs/2609.31647
#OpenSourceAI #DevTools #MLResearch
https://arxiv.org/abs/2609.31647
New paper reframes LLM harness design as multi-objective optimization across accuracy, safety, and token cost, showing a single-phase proposer beats two-phase approaches. Useful reminder that evaluation-time code shapes model…
#AI #LLM #AISafety #MLResearch
https://arxiv.org/abs/2609.30967
#AI #LLM #AISafety #MLResearch
https://arxiv.org/abs/2609.30967
September 28, 2026 at 8:01 PM
New paper reframes LLM harness design as multi-objective optimization across accuracy, safety, and token cost, showing a single-phase proposer beats two-phase approaches. Useful reminder that evaluation-time code shapes model…
#AI #LLM #AISafety #MLResearch
https://arxiv.org/abs/2609.30967
#AI #LLM #AISafety #MLResearch
https://arxiv.org/abs/2609.30967
DeepMind's new AI literally dreams about its past tries, turning failures into fresh insights. The breakthrough could reshape how we benchmark learning. Curious? Dive into the details! #DeepMind #DreamingAI #MLResearch
🔗 aidailypost.com/news/google-...
🔗 aidailypost.com/news/google-...
September 19, 2026 at 11:39 AM
DeepMind's new AI literally dreams about its past tries, turning failures into fresh insights. The breakthrough could reshape how we benchmark learning. Curious? Dive into the details! #DeepMind #DreamingAI #MLResearch
🔗 aidailypost.com/news/google-...
🔗 aidailypost.com/news/google-...