Enter your search text in the box above

Select a result to preview

Large Language Models, their capabilities, limitations, and theoretical foundations

表格 63 results
NameCategoriesTopicsReferencesCredibilityCreateDateUpdateDateTitleRelatedNotesNavigationOrderEleventyTemplateEngineOverrideCreated
20260414_gems_agent_native_multimodal_generation_memory_skillspaper_analysisagent_architecture, multimodal, memory_mechanism, long_term_memory, llmshanghai_ai_lab, sjtu100GEMS: Agent-Native Multimodal Generation with Memory and Skills[object Object][object Object]md
20260419_memskill_learning_evolving_memory_skillspaper_analysisself_evolving_agents, memory_mechanism, agent_architecture, llm, long_term_memorynanyang_technological_university, tsinghua_university, university_of_illinois_urbana_champaign100MemSkill: Learning and Evolving Memory Skills for Self-Evolving AgentsGEMS: Agent-Native Multimodal Generation with Memory and Skills[object Object][object Object]md
20260422_llm_semantic_reasoners_not_symbolic_reasonerspaper_analysisllm, reasoning, cognitive_science, language_philosophypku, bigai, proofwriter100Large Language Models are In-Context Semantic Reasoners rather than Symbolic ReasonersHippoRAG: Neurobiologically Inspired Long-Term Memory, M3-Agent: Multimodal Agent with Long-Term Memory, Memory OS of AI Agent[object Object][object Object]md
20260423_agentevolver_self_evolving_agentpaper_analysisagent_architecture, reinforce_learning, llm, reasoningtongyi_lab, appworld, bfcl_v3100AgentEvolver: Towards Efficient Self-Evolving Agent System[object Object][object Object]md
20260423_ai_reasoning_deep_learning_symbolic_neuralpaper_analysisreasoning, symbolic_reasoning, llm, knowledge_graph, neuro_science, cognitive_sciencebeihang_university100AI Reasoning in Deep Learning Era: From Symbolic AI to Neural Symbolic AIattentive_reasoning_queries[object Object][object Object]md
20260423_attentive_reasoning_queriespaper_analysisllm, agent_architecture, reasoning, cognitive_science, context_engineeringemcie_co100Attentive Reasoning Queries: Optimizing Instruction-Following in LLMsai_reasoning_deep_learning_symbolic_neural[object Object][object Object]md
20260423_meta_harness_end_to_end_optimizationpaper_analysiscontext_engineering, agent_architecture, llm, reasoning, ragstanford_iris_lab, terminalbench_2, krafton100Meta-Harness: End-to-End Optimization of Model Harnesses[object Object][object Object]md
20260424_context_engineering_2_overviewpaper_analysiscontext_engineering, llm, agent_architecture, reasoningsjtu_gair_lab, tongyi_lab100context_engineering_2_overview[object Object][object Object]md
20260424_llm_knowledge_graphs_opportunities_challengespaper_analysisknowledge_graph, llm, rag, reasoning100llm_knowledge_graphs_opportunities_challenges[object Object][object Object]md
20260425_chain_of_thought_prompting_elicits_reasoningpaper_analysisllm, reasoning, cognitive_science, context_engineering, multi_hop_reasoninggoogle_brain, chain_of_thought, gsm8k100Chain-of-Thought Prompting Elicits Reasoning in Large Language Models[object Object][object Object]md
20260425_reca_integrated_acceleration_cooperative_embodied_agentspaper_analysisllm, agent_architecture, memory_mechanism, embodied_ai, multi_hop_reasoninggeorgia_tech, harvard_university100ReCA: Integrated Acceleration for Real-Time and Efficient Cooperative Embodied Autonomous Agents[object Object][object Object]md
20260427_agentsquare_automatic_llm_agent_searchpaper_analysisagent_architecture, llm, reasoning, memory_mechanism, reinforce_learningtsinghua_fib_lab, neural_architecture_search100AgentSquare: Automatic LLM Agent Search in Modular Design Space[object Object][object Object]md
20260427_detecting_hallucinations_semantic_entropypaper_analysisllm, reasoning, agent_architecture, memory_mechanism, reinforce_learningoatml_oxford, neural_architecture_search100Detecting hallucinations in large language models using semantic entropy[object Object][object Object]md
20260428_rubrics_as_rewardspaper_analysisreinforce_learning, llm, reasoning, reward_modelingscale_ai95Rubrics as Rewards: Reinforcement Learning Beyond Verifiable Domains[object Object][object Object]md
20260429_agent_workflow_memorypaper_analysisagent_architecture, memory_mechanism, llm, gui_agent, long_term_memory, cognitive_sciencecmu, mit, mind2web, webarena100Agent Workflow Memorydisentangling_memory_reasoning_llm[object Object][object Object]md
20260429_disentangling_memory_reasoning_llmpaper_analysismemory_mechanism, reasoning, llm, multi_hop_reasoning, chain_of_thought, cognitive_sciencerutgers_university, ohio_state_university, ucsb, truthfulqa, strategyqa, commonsenseqa100Disentangling Memory and Reasoning Ability in Large Language Modelsagent_workflow_memory[object Object][object Object]md
20260430_catastrophic_forgetting_implicit_inferencepaper_analysiscatastrophic_forgetting, llm, memory_mechanism, reasoning, context_engineeringcmu, conjugate_prompting100Understanding Catastrophic Forgetting in Language Models via Implicit Inference[object Object][object Object]md
20260430_hallucination_survey_llmpaper_analysishallucination, llm, survey, rag, reinforce_learningharbin_institute_of_technology100A Survey on Hallucination in Large Language Models: Principles, Taxonomy, Challenges, and Open Questionscatastrophic_forgetting_implicit_inference[object Object][object Object]md
20260503_aflow_automating_agentic_workflow_generationpaper_analysisagent_architecture, llm, reasoning, workflow_optimizationdeepwisdom, monte_carlo_tree_search, claw100AFLOW: Automating Agentic Workflow Generation[object Object][object Object]md
20260506_openclaw_rl_train_any_agent_simply_by_talkingpaper_analysisreinforce_learning, agent_architecture, llm, self_evolving_agents, reasoninggen_verse100OpenClaw-RL: Train Any Agent Simply by Talking[object Object][object Object]md
20260507_scaling_large_language_model_multi_agent_collaborationpaper_analysisllm, agent_architecture, workflow_optimization, distributed_systemstsinghua_university, peng_cheng_laboratory100Scaling Large Language Model-based Multi-Agent Collaboration[object Object][object Object]md
20260508_memgpt_towards_llms_as_operating_systemspaper_analysisllm, memory_mechanism, agent_architecture, long_term_memory, context_engineeringuc_berkeley, memgpt100MemGPT: Towards LLMs as Operating Systems[object Object][object Object]md
20260511_rl_long_horizon_reasoning_llm_expressivenesspaper_analysisreinforce_learning, reasoning, llm, test_time_scaling, symbolic_reasoningpurdue_university, georgia_tech, scalelogic100Can RL Teach Long-Horizon Reasoning to LLMs? Expressiveness Is Keychain_of_thought_prompting_elicits_reasoning, reasoningbank_scaling_agent_self_evolving_reasoning_memory[object Object][object Object]md
20260512_gasim_graph_accelerated_hybrid_social_simulationpaper_analysisagent_architecture, memory_mechanism, knowledge_graph, llm, multi_agent_systemsuestc, graph_attention_network, mem0100GASim: A Graph-Accelerated Hybrid Framework for Social Simulationmemgpt_towards_llms_as_operating_systems[object Object][object Object]md
20260512_learning_fast_slow_llms_adapt_continuallypaper_analysiscatastrophic_forgetting, continual_learning, memory_mechanism, reinforce_learning, llmuc_berkeley100Learning, Fast and Slow: Towards LLMs That Adapt Continually20260504_reasoningbank_scaling_agent_self_evolving_reasoning_memory, 20260511_rl_long_horizon_reasoning_llm_expressiveness[object Object][object Object]md
20260512_meme_multi_entity_evolving_memory_evaluationpaper_analysismemory_mechanism, long_term_memory, agent_architecture, llm, reasoning, evaluationkaist_ai, tuebingen_ai_center, naver_ai_lab100MEME: Multi-entity & Evolving Memory Evaluationmemory_os_of_ai_agent, memgpt_towards_llms_as_operating_systems, disentangling_memory_reasoning_llm[object Object][object Object]md
20260515_apwa_parallelizable_agentic_workflowspaper_analysismulti_agent_systems, agent_architecture, distributed_systems, workflow_optimization, reasoning, llmnortheastern_university, apwa, ray, pii_300k, schemabench, summarybench95APWA: A Distributed Architecture for Parallelizable Agentic Workflowsscaling_large_language_model_multi_agent_collaboration, gasim_graph_accelerated_hybrid_social_simulation[object Object][object Object]md
20260517_heterogeneous_temporal_memory_governance_llm_personapaper_analysismemory_mechanism, llm, long_term_memory, rag, reasoning, persona_consistencyuestc, arpm100A Heterogeneous Temporal Memory Governance Framework for Long-Term LLM Persona Consistencymemgpt_towards_llms_as_operating_systems, meme_multi_entity_evolving_memory_evaluation[object Object][object Object]md
20260518_memlineage_lineage_guided_llm_agent_memorypaper_analysismemory_mechanism, agent_architecture, llm, agent_security, lineage_trackingiie_cas100MemLineage: Lineage-Guided Enforcement for LLM Agent Memory[object Object][object Object]md
20260520_amr_sd_token_level_credit_assignmentpaper_analysisreinforce_learning, reasoning, llm, reward_modeling, self_evolving_agentsgrpo, dapo, meituan, sciknoweval100AMR-SD: Asymmetric Meta-Reflective Self-Distillation for Token-Level Credit Assignmentrl_long_horizon_reasoning_llm_expressiveness, skillos_learning_skill_curation_self_evolving_agents[object Object][object Object]md
20260520_skillgenbench_benchmarking_skill_generation_llm_agentspaper_analysisagent_architecture, evaluation, skill_curation, llm, reasoningquanta_alpha, sjtu, pku, nus, xjtu, tsinghua_university, sufe, ntu, ucas100SkillGenBench: Benchmarking Skill Generation Pipelines for LLM Agentsontology_skill_analysis[object Object][object Object]md
20260521_methodology_selecting_composing_runtime_architecture_patterns_production_llm_agentspaper_analysisagent_architecture, distributed_systems, multi_agent_systems, llm, reasoningstanford_iris_lab, claw100A Methodology for Selecting and Composing Runtime Architecture Patterns for Production LLM Agentsagent_memory_and_reasoning_frontier_survey_2026, memgpt_towards_llms_as_operating_systems, memlineage_lineage_guided_llm_agent_memory[object Object][object Object]md
20260523_moss_self_evolution_source_level_rewritingpaper_analysisself_evolving_agents, agent_architecture, llm, multi_agent_systemsclaw, hkust_gz100MOSS: Self-Evolution through Source-Level Rewriting in Autonomous Agent Systemsmethodology_selecting_composing_runtime_architecture_patterns_production_llm_agents, skillos_learning_skill_curation_self_evolving_agents, agent_memory_and_reasoning_frontier_survey_2026[object Object][object Object]md
20260524_vector_policy_optimization_vpopaper_analysisreinforce_learning, llm, test_time_scaling, reasoning, reward_modelingmit, grpo100Vector Policy Optimization: Training for Diversity Improves Test-Time Search[object Object][object Object]md
20260525_gated_deltanet_2_decoupling_erase_write_linear_attentionpaper_analysismemory_mechanism, llm, reasoning, long_term_memory, reinforce_learningnvidia, deltanet100Gated DeltaNet-2: Decoupling Erase and Write in Linear Attention[object Object][object Object]md
20260528_unlocking_working_memory_latent_reasoningpaper_analysismemory_mechanism, reasoning, llm, reasoning_memory, context_engineeringlit_ai_lab, nxai_gmbh, gsm8k100Unlocking the Working Memory of Large Language Models for Latent Reasoningchain_of_thought_prompting_elicits_reasoning, memgpt_towards_llms_as_operating_systems, agent_memory_and_reasoning_frontier_survey_2026[object Object][object Object]md
20260529_stop_comparing_llm_agents_without_disclosing_harnesspaper_analysisagent_architecture, evaluation, llm, multi_agent_systems, reasoningtulane_university, rutgers_university, virginia_tech100Stop Comparing LLM Agents Without Disclosing the HarnessStop Comparing LLM Agents Without Disclosing the Harness[object Object][object Object]md
20260602_lintree_explicitly_structured_search_historiespaper_analysis|paper_analysisreasoning, llm, agent_architecture, memory_mechanismnus, oatml_oxford100LinTree: Improving LLM Reasoning with Explicitly Structured Search Historieschain_of_thought_prompting_elicits_reasoning[object Object][object Object]md
20260604_agent_memory_characterization_system_implicationspaper_analysismemory_mechanism, agent_architecture, long_term_memory, llm, evaluationstanford_university, ku_leuven, memory_agent_bench, memgpt, graphrag95Agent Memory: Characterization and System Implications of Stateful Long-Horizon Workloadsmemgpt_towards_llms_as_operating_systems, memlineage_lineage_guided_llm_agent_memory, calmem_dual_memory_conversational_ai, unlocking_working_memory_latent_reasoning, aip_graph_representation_for_learning_and_governing_agent_skills[object Object][object Object]md
20260604_language_models_need_sleep_self_modify_consolidate_memoriespaper_analysismemory_mechanism, self_evolving_agents, continual_learning, reinforce_learning, reasoning, llmgoogle_brain, knowledge_seeding, sleep_paradigm100Language Models Need Sleep: Learning to Self-Modify and Consolidate Memoriesmemory_os_of_ai_agent, hipporag_neurobiologically_inspired_long_term_memory, muse_autoskill_self_evolving_skill_memory[object Object][object Object]md
20260604_pretraining_recurrent_networks_without_recurrencepaper_analysismemory_mechanism, recurrent_neural_networks, llm, reinforce_learning, neuro_sciencemit, supervised_memory_training, backpropagation_through_time100Pretraining Recurrent Networks without Recurrence[object Object][object Object]md
20260611_trace_unified_rollout_budget_allocation_agentic_rlpaper_analysisreinforce_learning, reasoning, agent_architecture, test_time_scaling, llmtsinghua_university, hotpotqa, bfcl_v3, monte_carlo_tree_search100TRACE: A Unified Rollout Budget Allocation Framework for Efficient Agentic Reinforcement Learningrl_long_horizon_reasoning_llm_expressiveness, agent_memory_and_reasoning_frontier_survey_2026[object Object][object Object]md
20260617_evolvenav_proactive_preflection_self_evolving_memorypaper_analysisembodied_ai, semantic_navigation, self_evolving_agents, llm, memory_mechanismhkust_gz, nanyang_technological_university, habitat_simulator100EvolveNav: Proactive Preflection and Self-Evolving Memory for Zero-Shot Object Goal Navigation[object Object][object Object]md
20260618_fixed_point_reasoners_stable_adaptive_deep_looped_transformerspaper_analysisreasoning, llm, recurrent_neural_networks, memory_mechanism, agent_architecturetuebingen_ai_center, eth_zurich, liquidai, max_planck_institute100Fixed-Point Reasoners: Stable and Adaptive Deep Looped Transformers[object Object][object Object]md
20260622_how_transparent_is_diffusiongemmapaper_analysisllm, reasoning, agent_security, evaluation, interpretabilitygoogle_brain100How Transparent is DiffusionGemma?detecting_hallucinations_semantic_entropy[object Object][object Object]md
20260624_evoarena_tracking_memory_evolutionpaper_analysismemory_mechanism, agent_architecture, llm, evaluation, embodied_ainus, ntu, terminalbench_2100EvoArena: Tracking Memory Evolution for Robust LLM Agents in Dynamic EnvironmentsSelf-Harness: Harnesses That Improve Themselves[object Object][object Object]md
20260624_self_harnesspaper_analysisagent_architecture, self_evolving_agents, llm, reasoning, test_time_scalingshanghai_ai_lab, terminalbench_2100Self-Harness: Harnesses That Improve Themselves[object Object][object Object]md
20260626_progress_advantage_llm_agentspaper_analysisreinforce_learning, reward_modeling, test_time_scaling, agent_architecture, llm, reasoninguniversity_of_wisconsin_madison, argonne_national_laboratory100Neglected Free Lunch from Post-training: Progress Advantage for LLM Agentsrl_long_horizon_reasoning_llm_expressiveness, vector_policy_optimization_vpo, trace_unified_rollout_budget_allocation_agentic_rl[object Object][object Object]md
20260629_carve_content_aware_recurrent_value_efficiencypaper_analysismemory_mechanism, recurrent_neural_networks, llm, reasoning, linear_attentioncarve, deltanet100CARVE: Content-Aware Recurrent with Value Efficiency for Chunk-Parallel Linear AttentionGated DeltaNet-2: Decoupling Erase and Write in Linear Attention[object Object][object Object]md
20260629_memory_r1_enhancing_llm_agents_manage_utilize_memories_rlpaper_analysisagent_architecture, memory_mechanism, reinforce_learning, llm, long_term_memorylmu_munich, mem0, locomo_bench, grpo100Memory-R1: Enhancing Large Language Model Agents to Manage and Utilize Memories via Reinforcement Learningmemgpt_towards_llms_as_operating_systems, muse_autoskill_self_evolving_skill_memory, gems_agent_native_multimodal_generation_memory_skills, amr_sd_token_level_credit_assignment, agent_memory_characterization_system_implications[object Object][object Object]md
20260630_from_tokens_to_states_llms_world_modelspaper_analysisllm, reasoning, world_model, memory_mechanism, agent_architecturejepa, othello_gpt100From Tokens to States: LLMs as a Special Case of World Models and the Continuous Path Beyondmemory_os_of_ai_agent, rl_long_horizon_reasoning_llm_expressiveness[object Object][object Object]md
20260702_memory_in_the_age_of_ai_agents_a_surveypaper_analysismemory_mechanism, agent_architecture, survey, llm, long_term_memorynus, fudan_university, pku, renmin_university_of_china100Memory in the Age of AI Agents: A Survey — Forms, Functions and Dynamics[object Object][object Object]md
20260705_llm_agents_social_structure_latent_objective_emergencepaper_analysismulti_agent_systems, social_simulation, agent_architecture, reasoning, llm, evaluationcmu, dual_channel_debate, llm_agora100What LLM Agents Say When No One Is Watching: Social Structure and Latent Objective Emergence in Multi-Agent Debatesself_harness, contagion_networks_evaluator_bias_multi_agent_llm[object Object][object Object]md
20260706_demopsd_disagreement_modulated_policy_self_distillationpaper_analysisreinforce_learning, reasoning, llm, self_evolving_agentsgrpo, sciknoweval, kl_distillation, gpqa100DemoPSD: Disagreement-Modulated Policy Self-Distillationrl_long_horizon_reasoning, vector_policy_optimization[object Object][object Object]md
20260708_adacurl_adaptive_curriculum_rlpaper_analysisreinforce_learning, llm, reasoning, agent_architecture, evaluationamap100AdaCuRL: Adaptive Curriculum Reinforcement Learning with Invalid Sample Mitigation and Historical Revisiting[object Object][object Object]md
20260708_solving_million_step_llm_zero_errorspaper_analysismulti_agent_systems, reasoning, agent_architecture, llm, test_time_scalingcognizant_ai_lab, ut_austin100Solving a Million-Step LLM Task with Zero Errors[object Object][object Object]md
20260710_let_the_agent_search_tkgqapaper_analysisknowledge_graph, multi_hop_reasoning, llm, agent_architecture, reasoningat2qa, multi_tq, timeline_cronquestion, timeline_icews_actor100Let the Agent Search: Autonomous Exploration Beats Rigid Workflows in Temporal Question Answering[object Object][object Object]md
20260713_metacognition_in_llms_foundations_progress_opportunitiespaper_analysis, domain_surveyllm, reasoning, cognitive_science, memory_mechanism, agent_architectureyale_university, uc_irvine100Metacognition in LLMs: Foundations, Progress, and Opportunitieschain_of_thought_prompting_elicits_reasoning, ai_reasoning_deep_learning_symbolic_neural, memory_os_of_ai_agent[object Object][object Object]md
20260714_comprehensive_survey_knowledge_graph_reasoning_approaches_applicationspaper_analysisknowledge_graph, multi_hop_reasoning, reasoning, llm, survey100A Comprehensive Survey of Knowledge Graph Reasoning: Approaches and Applications[object Object][object Object]md
20260714_rubrics_as_rewards_reinforcement_learning_beyond_verifiable_domainspaper_analysisreinforce_learning, reward_modeling, llm, reasoning, evaluationscale_ai100Rubrics as Rewards: Reinforcement Learning Beyond Verifiable Domains[object Object][object Object]md
20260719_t2mlr_transformer_temporal_middle_layer_recurrencepaper_analysisreasoning, llm, memory_mechanism, recurrent_neural_networks, reasoning_memoryprinceton_university, chain_of_thought, backpropagation_through_time, gsm8k, hotpotqa100T²MLR: Transformer with Temporal Middle-Layer Recurrencesearchos_v1_open_domain_information_seeking_agent_collaboration, chain_of_thought_prompting_elicits_reasoning, memory_r1_enhancing_llm_agents_manage_utilize_memories_rl[object Object][object Object]md
20260723_copy_less_ground_more_evidence_aware_rlpaper_analysisreasoning, reinforce_learning, reward_modeling, context_engineering, llmpku, alibaba_group, gear, gspo, ruler_benchmark100Copy Less, Ground More: Overcoming Repetitive Copying in Long-Context Reasoning via Evidence-Aware Reinforcement Learning20260703_self_evolving_world_models_llm_agent_planning, 20260719_searchos_v1_open_domain_information_seeking_agent_collaboration[object Object][object Object]md
20260731_memrl_self_evolving_agents_runtime_rl_episodic_memorypaper_analysismemory_mechanism, reinforce_learning, agent_architecture, llm, reasoning, long_term_memorysjtu, xidian_university, nus, shanghai_innovation_institute, memtensor100MemRL: Self-Evolving Agents via Runtime Reinforcement Learning on Episodic Memory[object Object][object Object]md
Powered by Forestry.md