DiscoLoop: Looping Discrete Embeddings and Continuous Hidden States for Multi-hop Reasoning
Paper • 2607.00341 • Published
Hui
-
-
J-CoT: Chain-of-Thought in J-Space
Paper • 2607.21981 • Published • 2 -
LongHorizon-Harness: Advancing Long-Horizon Agents for Real-World Tasks
Paper • 2608.01964 • Published • 170 -
BDH-CQ: In-Context Learning with Recurrent Latent Reasoning
Paper • 2608.09888 • Published • 745 -
Spark-to-Paper: End-to-End Research Paper Generation as a Composable Skill
Paper • 2608.11924 • Published • 61 -
Agent Memory Distillation: Empowering Small LLM Agents with Hierarchical Teacher Memory
Paper • 2608.07169 • Published • 45 -
ToolSpace: Scaling Tool-Centric Environment Synthesis for Long-Horizon Agents
Paper • 2611.05036 • Published • 43 -
TAC-CoT: Tool-Empowered Implicit Chain-of-Thought via Interleaving Explicit Tokens
Paper • 2611.01926 • Published • 17 -
Latent Sparse Attention: Towards Subquadratic Ultra-Long Context Attention
Paper • 2610.42147 • Published • 3 -
Dynamic Linear Attention
Paper • 2606.10650 • Published • 9 -
Large Flow Language Models
Paper • 2610.28559 • Published • 16 -
NCP-ArchPreview Technical Report: Moving towards Latent Space Language Models through Next Concept Prediction
Paper • 2609.10715 • Published • 167 -
Unlocking Lossless Speedups in LLMs via Discrete Diffusion
Paper • 2609.04010 • Published • 111 -
Muir: Linear Attention with Sensitivity-Allocated Gating and Matched-Filter Readout
Paper • 2611.08376 • Published • 14 -
MoME: Mixture-of-Memory Embeddings for Context-Aware Sparse Lookup
Paper • 2609.15126 • Published • 10 -
OpenGCUA: Toward a Fully Open Stack for Generalist Computer Use
Paper • 2610.22200 • Published • 71 -
Harness^2: Recursive Agent Harnessing for an Open World
Paper • 2610.14224 • Published • 70 -
Hybrid Computer Use: A Unified Programmatic, Semantic, and Visual Interface for General Agents
Paper • 2610.41432 • Published • 19 -
D3D: Decoupled Diffusion Drafting for Asynchronous Speculative Decoding
Paper • 2611.06425 • Published • 20 -
Pack Light, CarryOn: Building Object-Oriented Context for Long-Horizon Agents
Paper • 2610.18784 • Published • 118 -
Latent Reasoning in Continuous Space for Unified Multimodal Models
Paper • 2610.42149 • Published • 44 -
Hierarchical Continuous Diffusion Language Models
Paper • 2610.02193 • Published • 71 -
Lngram: N-gram Conditional Memory in Latent Space
Paper • 2605.24869 • Published • 2 -
Library of Layers (LoL): Adaptively Ordering Transformer Layers at Runtime
Paper • 2611.03526 • Published • 12 -
Virtual Head Attention
Paper • 2610.17959 • Published • 17 -
Kronecker Delta Rule for Long Context Linear Attention
Paper • 2610.26834 • Published • 31 -
Gated Recurrent Transformers: Expressive Depth through Recurrent Modulation
Paper • 2608.15062 • Published • 12 -
Copy-Paste Language Models
Paper • 2610.19247 • Published • 9 -
The COUNT: Counting N-grams Unlocks Scaling on Host Memory and Training-Free Adaptation
Paper • 2610.20456 • Published • 7 -
FactorEngram: Factorized N-gram Memory with Basis-Level Gating for Language Models
Paper • 2609.35578 • Published • 4 -
Think Less, Locally: Local Preference Learning for Reasoning Compression
Paper • 2611.04814 • Published • 66 -
Latent Reasoning with Normalizing Flows
Paper • 2606.06447 • Published • 8 -
Formalizing Latent Thoughts: Four Axioms of Thought Representation in LLMs
Paper • 2606.27378 • Published • 49 -
Second Thought: Reasoning in Parallel as LLM Agents Act and Observe
Paper • 2608.13667 • Published • 17 -
Thought-Level Beam Search for Reasoning
Paper • 2608.08020 • Published • 8 -
DiffusionBlocks: Blockwise Training for Generative Models via Score-Based Diffusion
Paper • 2506.14202 • Published • 10 -
LatentPress: Context Compression Beyond Text and Vision
Paper • 2609.01507 • Published • 24 -
Real Long-Term Memory for AI: A 50-Million-Token Window That Is Faster and Cheaper Than Recompute
Paper • 2610.10845 • Published • 10 -
A*-Thought-V2: Efficient Latent Reasoning via Geometric Dynamics of LLM
Paper • 2609.07821 • Published • 7 -
SpectralShift: Effective Context Window Extension of Gated DeltaNet via Spectral Reparameterization
Paper • 2609.14320 • Published • 6 -
What Matters for Latent Reasoning with Flow Matching
Paper • 2610.06666 • Published • 9 -
Periscope: Extending Frozen Language Models Beyond Their Context Window
Paper • 2610.04047 • Published • 5 -
Learning to Steer, Steering to See: Unveiling the Geometry of RLVR in Large Language Models via Trainable Vectors
Paper • 2609.34344 • Published • 5 -
HyperThink: Text-to-Parameter Hypernetworks for Efficient Reasoning
Paper • 2610.03039 • Published • 6 -
Understanding Evolution Strategies for LLM Reasoning: Broader Reasoning Coverage than GRPO
Paper • 2608.27351 • Published • 19 -
When2Think: Learning When and How Much to Reason
Paper • 2609.19671 • Published • 20 -
Hierarchical Residuals: A Minimalist Decoupling of Residual Width from Compute Width
Paper • 2610.41711 • Published • 9 -
LowRAM: Low-Rank Keys and Amortised Values for Composable KV Cache Compression at Test-Time
Paper • 2611.07025 • Published • 36