ICLR 2026PastMath & reasoning
Workshop on Latent & Implicit Thinking – Going Beyond CoT Reasoning
LIT Workshop @ ICLR 2026
- Submission deadline
- Feb 9, 2026, 12:00 UTCOpenReview-synced 2026-02-09 12:00 UTC (as of 2026-06-23) — extensions on OpenReview are applied automatically; verify on the website.
- Submission portal
- OpenReview
- Notes
- Topics were auto-suggested and may be imprecise — edits welcome.
Accepted papers (78)
Fetched from OpenReview (v2) on 2026-06-10.
ActivationReasoning: Logical Reasoning in Latent Activation Spaces
Adaptive Loops and Memory in Transformers: Think Harder or Know More?
All Roads Lead to Rome: Distilling Verifiable Reasoning via Shared Decision Pivots
Analyzing and Improving Chain-of-Thought Monitorability Through Information Theory
Are Latent Reasoning Models Easily Interpretable?
Beyond Memorization: Extending Reasoning Depth with Recurrence, Memory and Test-Time Compute Scaling
Breaking the Reversal Curse in Autoregressive Language Models via Identity Bridge
Bypassing the Rationale: Causal Auditing of Implicit Reasoning in Language Models
Can the Future Inform the Present? Investigating Latent Lookahead Refinement via Multi-Token Prediction
ConFu: Contemplate the Future for Better Speculative Sampling
Cross-Layer Clustering for Stochastic Parameter Decomposition
Debugging code world models
Denoising is not the End: Discrete Diffusion Language Models with Self-Correction
Discovering Interpretable Algorithms by Decompiling Transformers to RASP
Do Depth-Grown Models Overcome the Curse of Depth? An In-Depth Analysis
Dual-Channel Steering: Combining Explicit Prompting and Implicit Parameter Modulation for Reasoning Diversity
Dynamic Large Concept Models: Latent Reasoning in an Adaptive Semantic Space
Dynamics Within Latent Chain-of-Thought: An Empirical Study of Causal Structure
Emergent Analogy in Transformers
Encode, Think, Decode: Scaling test-time reasoning with recursive latent thoughts
Energy-Conditioned Thinking: A Three-State Framework for Adaptive Depth and Halting
From Growing to Looping: A Unified View of Iterative Computation in LLMs
How Do Latent Reasoning Methods Perform Under Weak and Strong Supervision?
How to Train Your HRM
Implicit Statistical Inference in Transformers: Approximating Likelihood-Ratio Tests In-Context
Inference-Time Rethinking with Latent Thought Vectors for Math Reasoning
Is continuous CoT better suited for multilingual reasoning?
LaneRoPE: Positional Encoding for Collaborative Parallel Reasoning and Generation
LASER: Low-Rank Activation SVD for Efficient Recursion
Latent-DARM: Bridging Discrete Diffusion And Autoregressive Models For Reasoning
LatentChem: From Textual CoT to Latent Thinking in Chemical Reasoning
Learning Efficient Latent Reasoning with Abstract Chain-of-Thought
Learning from Partial Chain-of-Thought via Truncated-Reasoning Self-Distillation
Learning Multi-step Reasoning via Persistent Latent State Propagation
Learning State-Tracking from Code: REPL Traces and Probabilistic Automata
Learning to Execute Graph Algorithms Exactly with Graph Neural Networks
Lightweight Latent Reasoning for Narrative Tasks
LLMs Encode Their Failures: Predicting Success from Pre-Generation Activations
LOOK BEFORE YOU LEAP: THERMODYNAMIC ARBITRATION OF PARAMETRIC AND NON-PARAMETRIC KNOWLEDGE IN LLM AGENTS VIA SELF-REGULATING MEMORY ARCHITECTURES
Mechanisms of Introspective Awareness
Mechanistic Analysis Of Universality: Numerical Comparison Circuits Across Transformer Architectures
Mechanistic Evidence for Faithfulness Decay in Chain-of-Thought Reasoning
MILR: Improving Multimodal Image Generation via Test-Time Latent Reasoning
Modeling Tool Use in Transformers via Computation Oracles
Offline RL with Hierarchical Action Chunking
On the Residual Scaling of Looped Transformers: Stability and Transferability
One Step Forward and K Steps Back: Better Reasoning with Denoising Recursion Models
Parcae: A Dynamical Systems Perspective to Stable Looped LLMs
Polestar-Cache: Reconciling Parallel Decoding and Accuracy in Diffusion LLMs via Token Drift-Aware KV Cache Recalibration
Reasoning Within the Mind: Dynamic Multimodal Interleaving in Latent Space
RECURRENT-DEPTH VLA: IMPLICIT TEST-TIME COMPUTE SCALING OF VISION–LANGUAGE–ACTION MODELS VIA LATENT ITERATIVE REASONING
Recursive Reasoning as Attractor Landscape Search: Mechanistic Dynamics of the Tiny Recursive Model
Rejection Mixing: Fast Semantic Propagation of Mask Tokens for Efficient DLLM Inference
RFG: Test-Time Scaling for Diffusion Large Language Model Reasoning with Reward-Free Guidance
SEMIE: Semantic Entropy-Informed Decoding
Single-Position Intervention Fails: Distributed Output Templates Drive In-Context Learning
T2MLR: Transformer with Temporal Middle-Layer Recurrence
Task-Specific Knowledge Distillation via Intermediate Probes
Test-Time Meta-Adaptation with Self-Synthesis
The Illusion of Superposition in Latent CoT via Soft Thinking
The Mechanistic Invariance Test: Genomic Language Models Fail To Learn Positional Regulatory Logic
The Power of Power Law: Asymmetry Enables Compositional Reasoning
THINK DEEP, SPEAK ONCE: RELIT, A RECURSIVE LATENT IMPLICIT TRANSFORMER FRAMEWORK
Think-at-Hard: Selective Latent Iterations to Improve Reasoning Language Models
Thinking in Latents: Adaptive Anchor Refinement for Implicit Reasoning in LLMs
Thinking into the Future: Latent Lookahead Training for Transformers
Tiny Autoregressive Recursive Models
Tiny Recursive Reasoning with Mamba-2 Attention Hybrid
Transformers Provably Learn to Internalize Chain-of-Thought
TSLM: Tree-Structured Language Modeling for Divergent Thinking
Ulterior Motives: Detecting Misaligned Reasoning in Continuous Thought Models
Variational Latent Reasoning Guided by Rendered Chain-of-Thought
When does Chain-of-Thought Help: A Markovian Perspective
When Intermediate Supervision Doesn’t Help: Evidence from Recurrent CNNs
When Pruning Breaks Reasoning: Chain-of-Thought Similarity and Faithfulness in Language Models
When Shallow Wins: Silent Failures and the Depth–Accuracy Paradox in Latent Reasoning
Which Heads Matter for Reasoning? RL-Guided KV Cache Compression
ε-Leaf Enumeration: Non-Repeating Self-Consistency via Truncated Tree Search