CommunityMar 8, 20262026-03-085 minVerified
Combining LLM Agents With Science Models for Reliable Loops
A practical pattern: LLMs handle planning and interpretation, while science models provide constraint-based scoring and stopping gates.
What this topic covers
Signals, research, and debates around general intelligence and superintelligence.
Hub content is updated incrementally.
All Articles
SourceMar 8, 20262026-03-085 minVerified
Estimating Rankings via Pairwise LLM Comparisons and MCMC
Instead of long one-shot rankings, use pairwise LLM judgments and Bradley–Terry with Bayesian MCMC to estimate ranks and uncertainty.
SourceMar 8, 20262026-03-084 minVerified
Learnable Pixel Weights For Controllable Medical Diffusion Pipelines
Summarizes LAW: learnable per-pixel loss reweighting to address spatial imbalance in medical diffusion and segmentation, improving FID.
CommunityMar 8, 20262026-03-085 minVerified
Why Custom Instructions Personas Drift Under Hierarchy
Model Spec’s chain of command can override custom instructions, causing persona and reasoning drift. Design priorities, exceptions, and fallbacks to improve reproducibility.
SourceMar 7, 20262026-03-074 minVerified
Gating Robot Autonomy Using Deep Perception Uncertainty Signals
SPIRIT uses deep perception uncertainty to gate shared autonomy, switching between semi-autonomous manipulation and haptic teleoperation.
SourceMar 7, 20262026-03-074 minVerified
LegalBench And Auditable Argumentation For Legal LLMs
How LegalBench evaluates legal LLM reasoning beyond accuracy, emphasizing justification and auditability through structured argumentation and governance.
SourceMar 7, 20262026-03-074 minVerified
Memory Admission Control for Reliable LLM Agents
A practical look at memory admission control for LLM agents, reducing long-term memory pollution while improving auditability and metrics.
SourceMar 7, 20262026-03-074 minVerified
Multimodal Clinical Reasoning Needs Controlled Evaluation, Not Scores
In multimodal clinical reasoning, reported gains don’t guarantee safety; prioritize controlled evaluation, grounding, and auditable failure modes.
SourceMar 6, 20262026-03-063 minVerified
Cryo-SWAN Brings Voxel Density Maps Into 3D VAE
Cryo-SWAN is a voxel density-map VAE, reporting consistent reconstruction-quality gains across ModelNet40, BuildingNet, and ProteinNet3D.
CommunityMar 6, 20262026-03-064 minVerified
Designing AI Coding Quota Markets Under Terms Constraints
If/Then guide to AI coding quota marketplaces: structure roles, avoid key-transfer violations, and add SSDF-style verification.
SourceMar 6, 20262026-03-064 minVerified
Designing Terminal-Native Coding Agents With Context Pipelines
A shift from IDE plugins to terminal-native CLI coding agents, highlighting AGENTS.md and context pipelines that shape reliability and verification loops.
SourceMar 6, 20262026-03-064 minVerified
Extreme 2-Bit Quantization Can Break LLM Generation
Study compares six post-training 2-bit methods on a Polish 11B LLM, highlighting gaps between benchmarks and generation stability.
SourceMar 6, 20262026-03-065 minVerified
Vehicle Anchors Recover Metric Scale in GPS-Denied UAV Video
VANGUARD estimates GSD from monocular UAV video using small vehicles as anchors to recover metric scale without GPS or telemetry.
SourceMar 5, 20262026-03-054 minVerified
Chain-of-Thought Perturbations Reveal Hidden Fragility in Reasoning
CoT perturbations can sharply reduce accuracy. Unit conversion remains hard at scale; isolate checks and use self-consistency.
CommunityMar 5, 20262026-03-055 minVerified
ChatGPT Model Retirement Reshapes Tone, Safety, Creativity Balance
Retiring legacy ChatGPT models may shift tone, refusals, and creativity, reshaping the balance between expression and safety guardrails.
SourceMar 5, 20262026-03-054 minVerified
Consensus-Anchored Diffusion for Uncertain 3D Lesion Segmentation
Examines multi-rater 3D lesion segmentation, limits of vanilla diffusion, and VDD anchored to consensus priors improving GED/CI.
CommunityMar 5, 20262026-03-055 minVerified
Difficulty Illusions In LLMs And Evaluation Design
How LLMs create difficulty illusions, and how to design evaluation gates with scenarios, protocols, and multi-metric reporting.
SourceMar 5, 20262026-03-054 minVerified
GIPO Reuses Off-Policy Data With Gaussian Trust Weighting
GIPO targets scarce, stale interaction data by replacing hard importance-ratio clipping with log-ratio Gaussian trust weights for stable reuse.
SourceMar 5, 20262026-03-054 minVerified
Governance For Reliable Agentic AI In WebGIS Development
Reframes agentic AI failures as governance issues, proposing dual-helix governance with a Knowledge/Behavior/Skills architecture.
SourceMar 5, 20262026-03-055 minVerified
LLM-Guided Belief Shaping for Partially Observable TAMP
How LLM signals can shape belief in partially observable TAMP, and why calibration, uncertainty, and safety filters matter for reliability.
SourceMar 5, 20262026-03-054 minVerified
Managerial AI Advice Under Ambiguity and Sycophancy Risks
How ambiguity detection, clarification, and sycophancy control shape managerial AI advice quality, risk, and evaluation metrics.
SourceMar 5, 20262026-03-054 minVerified
MASS Enables Test-Time Self-Update Through Synthetic Data
MASS trains LLMs to synthesize per-problem data and self-update at test time, raising auditability, integrity, and reproducibility needs.
SourceMar 5, 20262026-03-054 minVerified
PlugMem Plugin Memory Cuts Context Bloat, Adds Risk
PlugMem externalizes long-term memory as a plug-in to reduce retrieval bloat and relevance loss, while highlighting persistent injection risks.
CommunityMar 5, 20262026-03-054 minVerified
Reproducible Visual Puzzle Evaluation Without Tool Leakage
Tool-free visual puzzle claims depend on fixed constraints: lock tools, image preprocessing, prompts, and logs for reproducibility.