Search papers, labs, and topics across Lattice.
Over 40% improvement in analytical efficiency could revolutionize how researchers conduct trajectory inference in single-cell transcriptomics.
Single-rollout sampling can dramatically improve the stability and effectiveness of RL training for large language models, outperforming traditional methods by a significant margin.
Current LLMs falter in complex deliberative collaboration tasks, revealing critical gaps in their reasoning capabilities even when aided by external tools.
AgentTether repairs over 65% of failures in complex LLM tasks without modifying the agent, revolutionizing how we ensure reliability in AI deployments.
Cortex outperforms traditional models by enabling zero-shot execution of complex long-horizon tasks, bridging the gap between high-level planning and low-level execution.
CompactionRL enables LLMs to effectively manage long-horizon tasks by summarizing context, leading to substantial performance gains in coding benchmarks.
UI-MOPD achieves a remarkable balance between retaining existing capabilities and adapting to new platforms, with task success rates that challenge conventional approaches in GUI agent learning.
State-of-the-art planners falter in long-tail scenarios, revealing critical gaps in autonomous driving safety and effectiveness.
ACE achieves a remarkable 70% success rate in constraint retrieval tasks without any task-specific retraining, showcasing the power of zero-shot workflow reasoning in robotic manipulation.
AgenticDataBench reveals that LLM-based data agents can be rigorously evaluated across diverse real-world scenarios, highlighting their strengths and weaknesses in handling complex data tasks.
Visualizing code dependencies can dramatically enhance issue-resolution performance, outperforming traditional text-based navigation methods.
GPT-5.5 not only tops the leaderboard in policy evolution but also reveals critical insights into how agents can optimize performance through strategic feedback utilization.
Runtime diagnoses from multi-faceted bug reproduction tests can significantly boost patch generation effectiveness, leading to a 75.7% resolution rate on verified issues.
Self-evolving agents could revolutionize enterprise AI by enabling continuous learning from real-world interactions, but current systems are falling short.
TACO reveals that agentic models can learn to optimize tool usage without external judges, achieving higher accuracy and efficiency in multimodal tasks.
Off-policy distillation fails in multi-task settings, but a two-phase approach combining it with on-policy refinement can achieve single-task expert performance across multiple tasks.
Even the most advanced LLMs struggle with consistent rubric verification, revealing substantial noise in scoring outputs across complex agentic scenarios.
Effective keyframe extraction can boost MLLM performance by over 2% on complex video-guided tasks, revealing a critical link between video understanding and procedural learning.
ProMSA achieves superior accuracy in KB-VQA by dynamically selecting retrieval strategies, outperforming traditional fixed pipelines.
OPID achieves a remarkable boost in agent performance by leveraging hierarchical skills extracted from on-policy trajectories, transforming sparse rewards into dense, actionable insights.
PhysEditWorld reveals that explicit control over physical parameters can transform how game world models interact with their environments, leading to more realistic and manipulable simulations.
EGG achieves a remarkable 2.13x speedup in GPU kernel generation, setting a new benchmark for performance in automated optimization.
ASSCG cuts inference latency by 60% while boosting performance scores in autonomous driving systems, redefining how LLMs can be efficiently integrated into fast-slow planning architectures.
Evolving hardware-aware compression techniques can outperform human designs, achieving unprecedented efficiency in deploying massive AI models.
Agents can now escape the Self-Confirmation Trap, leading to more reliable experience learning and improved self-evolution.
ReMMD-Agent achieves a remarkable 41.80% accuracy in detecting misinformation across complex multilingual and multi-image scenarios while slashing verification costs by up to 80%.
A unified runtime boundary and time-aware execution can boost LLM agent accuracy by over 2% in long-horizon tasks, revealing a critical leverage point for enhancing agent stability.
No single memory architecture is best for all tasks; performance hinges on how well memory structures align with specific workload challenges.
AOHP redefines how AI agents interact with operating systems, achieving a 21% boost in task completion and a dramatic cut in execution costs.
CUAs can achieve a 73.7% success rate on complex macOS tasks, but the secret to their performance lies in skill libraries, not just framework design.
H-RePlan achieves a remarkable increase in task completion rates by intelligently distinguishing between local and global recovery strategies in multi-device environments.
MobileForge achieves a remarkable 77.6% Pass@3 score for mobile GUI agents using only annotation-free adaptation data, setting a new standard in the field.
Closing the supervision gap in GUI agents boosts success rates from the low-30% range to over 50% through innovative skill-guided learning.
By harnessing implicit supervision from environment dynamics, EnvRL boosts RL success rates by over 4% on long-horizon tasks, revealing a new frontier in agentic learning.
WEQA achieves a 24% accuracy boost in wearable health question answering by dynamically adapting to the complexities of sensor data and user queries.
Cordon reveals that a transactional approach to LLM agent runtimes can drastically reduce irreversible failures while enhancing task integrity across complex workflows.
Semantic acceptance rates can be misleading, with up to 44.2% of models failing to prevent observable harm even when they pass initial checks.
Collective Skill Tree Search transforms LLMs into versatile agents capable of mastering complex tasks through a structured skill tree that enhances their adaptability and performance.
HABC achieves up to 92% success in complex robotic tasks by intelligently balancing viability and efficiency in sparse outcome scenarios.
Effective personalization in presentation generation hinges on a novel memory architecture that distinguishes between user profiles, session constraints, and execution history.
TrustedARI reduces communication overhead by over 39% while enabling secure, verifiable interactions between AI agents and external services.
Reliable phone automation hinges on mixed-action capabilities, with agents achieving a 75% success rate in real-world workflows.
LLMs are evolving from reactive chatbots to proactive digital colleagues, fundamentally changing how AI can assist in complex tasks.
PERIA-8B not only surpasses state-of-the-art models in spatial reasoning but does so with a fraction of their size, revealing the power of tool-augmented approaches.
Environment engineering, not just agent workflows, is the key to unlocking the full potential of autonomous scientific discovery, as demonstrated by EurekAgent's record-breaking results.
Adapter design can make or break coding performance in OpenClaw-style agents, with a full adapter boosting success rates by over 50 percentage points.
AutoPDE reveals that explicitly representing solver strategies can boost PDE solving reliability by over 14% compared to existing methods.
Infini Memory redefines long-term memory for LLMs, achieving a 64.7% score on MemoryAgentBench by structuring memory as topic documents that evolve over time.
Current autonomous AI agents are alarmingly unprepared for real-world adversarial attacks, often missing critical vulnerabilities in dynamic environments.
GUI-AC transforms how GUI agents learn by stabilizing their performance amidst the chaos of non-stationary environments.
