Research Feed Page 12

Search source-linked summaries of recent AI and machine-learning papers by topic, by date, or by whether they include code or a diagram.

Filter papers All papers

Browse by date

Resource filters

Sort options

Research results

Agents / Efficiency & Inference By Wei Lin 2026-08-20
The Third Restructuring of Software Form: From the Three-Tier Architecture to Storage, Models, and Agents

The authors propose Software 3.0, a new architecture that replaces traditional three-tier systems with a converged structure consisting of a persistent storage layer, a probabilistic intelligence core, and an agent-based execution loop.

Agents / Training & Fine-Tuning By Fengqing Jiang 2026-08-20
MidTool: Mid-training Data Synthesis for Agentic Tool Use

The researchers created a 20.3B-token corpus called MidTool-Mix to improve agentic tool-use capabilities in models during the mid-training phase rather than relying solely on post-training.

Agents / Benchmarks & Evals By Chengsong Huang 2026-08-20 201
EnvHarness: Awakening Static Worlds for Agent Learning

EnvHarness and EnvRigger dynamically modify static environments to improve LLM agent training efficiency and performance.

Agents / Safety & Alignment By Seongjae Kang 2026-08-20 2
PolicyGuide: From Guarding One Action to Guiding the Whole Workflow for Policy-Compliant LLM Agents

PolicyGuide introduces a workflow-based verification system that uses an external runtime graph to enforce organizational compliance in LLM agents.

Efficiency & Inference / Benchmarks & Evals By Adam Fisch 2026-08-20
Pandora's AI Model Routing Box: Efficient Allocation with Costly Value Estimation

The paper introduces a routing framework that reduces computational costs by intelligently deciding when to pay for accurate model value estimates rather than using cheaper, noisier alternatives.

Benchmarks & Evals / Agents By Zhipeng Xu 2026-08-20 37
SWE-bench Science: Can Coding Agents Resolve Engineering Tasks in Science?

The paper introduces SWE bench Science to evaluate coding agents on repository-level scientific software engineering tasks and analyzes how scientific guidance impacts their performance.

Agents / Efficiency & Inference By Yu Chen 2026-08-20
Optimal Skill Selection for LLM Agents with Provable Bicriteria Guarantees

The paper introduces a new algorithm called Best Prefix Selection to choose the most efficient set of skills for LLM agents within a fixed token budget.

Training & Fine-Tuning / Benchmarks & Evals By Qian Kou 2026-08-20
Inject, Align, Recover: Staged Post-Training for Retrieval-Free Document Knowledge Internalization

The researchers developed a staged training method called IAR to improve how models store and answer questions about specific document sets without needing retrieval systems.

Agents / Benchmarks & Evals By Kou Shi 2026-08-19 2
FACET: Preserving Source Intent and Executable State in Terminal Task Synthesis

The paper introduces FACET, a framework for generating consistent, executable terminal tasks by grounding task artifacts like instructions and verifiers in a shared containerized state.

Agents / Benchmarks & Evals By Yajing Bai 2026-08-18 8
HarnessRisk: A Lifecycle-Oriented Benchmark for Agent Harness Safety

HarnessRisk provides a lifecycle based framework for evaluating security vulnerabilities across six distinct operational phases in agentic systems.

Agents / Reinforcement Learning By Bo Liu 2026-08-19
SPADE: Self-Play in Adaptive Synthetic Executable Environments

SPADE improves agent performance by using an automated system that co-evolves training environments and reasoning agents through a continuous reinforcement learning loop.

Benchmarks & Evals / Multimodal By Keyu Tu 2026-08-18 141
SemComp-Bench: Benchmarking Semantic Task Completion in Video Generation

The paper introduces a dataset and evaluation protocol to measure how accurately video generation models can complete specific instructed outcomes while maintaining semantic grounding.

Agents / Reinforcement Learning By Yiming Du 2026-08-18 17
LEGO-RL: Harness-Native Reinforcement Learning for Coding Agents

Lego-RL is a framework that aligns native coding execution harnesses with policy-gradient training to improve agent performance and stability.

Agents / Efficiency & Inference By Alizer Wong 2026-08-19
Eureka: Task-Conditioned Meta-Agent Orchestration for Scientific Discovery

Eureka introduces a meta-agent architecture that dynamically promotes specialized agents to solve long-horizon scientific tasks while minimizing computational overhead.

Efficiency & Inference / Benchmarks & Evals By Genghan Zhang 2026-08-18 2
PTXBench: Benchmark and Adapt LLMs for GPU Kernel Optimization with Architecture-specific PTX

The authors introduce PTXBench and a supervised fine-tuning method to help LLMs write efficient architecture-specific GPU code.

Reasoning / Efficiency & Inference By Wei Yu 2026-08-19
Training-Free Inference-Time Self-Reflection and Cost-Bounded Early Stopping for Large Language Models

The paper introduces a training-free inference-time protocol that uses a self-critique loop and a confirmed sentinel to improve reasoning accuracy while early-stopping redundant computations.

Agents / Benchmarks & Evals By Silin Chen 2026-08-19
SkillForge: Self-Distilling Agents for Project-Specific Issue Resolution

SkillForge enhances software engineering agents by distilling repository-specific knowledge into reusable skills to solve project-specific issues.

Agents / Reinforcement Learning By Qingyao Li 2026-08-19
SkillGate: Training In-Policy Skill Selection in Long-Horizon Agents

SkillGate optimizes agent performance by partitioning training signals to stop task outcomes from interfering with how agents select procedural skills.

Agents / Safety & Alignment By Zonghao Ying 2026-08-18 5
Security Assessment of DeepSeek Harness with A.I.G: Evaluating Resistance to Indirect Prompt Injection

The paper uses an automated framework to evaluate how untrusted external data can manipulate agents in the DeepSeek Harness framework into performing unintended actions.

Agents / Benchmarks & Evals By Yanlun Tu 2026-08-19 43
SemaPLC: A Project-Grounded, Verification-Gated Agent Harness for PLC Code Generation

SemaPLC uses an agent-based workflow and multi-stage verification to ensure that LLM-generated PLC programs integrate correctly and execute reliably within existing industrial projects.