ConPress: Learning Efficient Reasoning from Multi-Question Contextual Pressure

• URL: http://arxiv.org/abs/2602.01472v1
• Date: Sun, 01 Feb 2026 22:31:19 GMT
• Title: ConPress: Learning Efficient Reasoning from Multi-Question Contextual Pressure
• Abstract summary: We identify a reproducible inference-time phenomenon, termed Self-Compression.<n>When multiple independent and answerable questions are presented within a single prompt, the model spontaneously produces shorter reasoning traces for each question.<n>We propose ConPress, a lightweight self-supervised fine-tuning approach.
• Score: 6.494457510860719
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Large reasoning models (LRMs) typically solve reasoning-intensive tasks by generating long chain-of-thought (CoT) traces, leading to substantial inference overhead. We identify a reproducible inference-time phenomenon, termed Self-Compression: when multiple independent and answerable questions are presented within a single prompt, the model spontaneously produces shorter reasoning traces for each question. This phenomenon arises from multi-question contextual pressure during generation and consistently manifests across models and benchmarks. Building on this observation, we propose ConPress (Learning from Contextual Pressure), a lightweight self-supervised fine-tuning approach. ConPress constructs multi-question prompts to induce self-compression, samples the resulting model outputs, and parses and filters per-question traces to obtain concise yet correct reasoning trajectories. These trajectories are directly used for supervised fine-tuning, internalizing compressed reasoning behavior in single-question settings without external teachers, manual pruning, or reinforcement learning. With only 8k fine-tuning examples, ConPress reduces reasoning token usage by 59% on MATH500 and 33% on AIME25, while maintaining competitive accuracy.

Related papers

• DeltaPrompts: Escaping the Zero-Delta Trap in Multimodal Distillation [49.98710755440242]
  Distillation enables compact Vision-Language Models (VLMs) to obtain strong reasoning capabilities.<n>We reveal a critical inefficiency in this approach: up to 69% of the prompts in standard chart / document reasoning datasets are effectively zero-delta.<n>We propose a staged synthesis pipeline that repurposes existing datasets as seeds, actively targeting student failure modes to produce better prompts.
  arXiv  Detail & Related papers  (2026-05-15T02:04:12Z)
• Reasoning Compression with Mixed-Policy Distillation [20.7258653384411]
  Reasoning-centric large language models often incur excessive token usage and high inference-time decoding cost.<n>We propose Mixed-Policy Distillation (MPD), a reasoning compression framework that transfers concise reasoning behavior from a larger-sized teacher to a smaller student.<n>MPD reduces token usage by up to 27.1% while improving performance across multiple reasoning benchmarks.
  arXiv  Detail & Related papers  (2026-05-09T08:04:47Z)
• Early Stopping for Large Reasoning Models via Confidence Dynamics [55.67938134245981]
  Large reasoning models rely on long chain-of-thought generation to solve complex problems.<n>Key challenge is determining when the model should stop reasoning and produce the final answer.<n>We propose CoDE-Stop, an early stopping method that leverages the dynamics of intermediate answer confidence to decide when to terminate reasoning.
  arXiv  Detail & Related papers  (2026-04-06T17:59:45Z)
• PLUME: Latent Reasoning Based Universal Multimodal Embedding [52.35354073629127]
  Universal multimodal embedding (UME) maps heterogeneous inputs into a shared retrieval space with a single model.<n>Recent approaches improve UME by generating explicit chain-of-thought (CoT) rationales before extracting embeddings.<n>We propose PLUME, a latent reasoning framework that advances UME by replacing verbalized CoT with a short autoregressive rollout of continuous latent states.
  arXiv  Detail & Related papers  (2026-04-02T14:04:53Z)
• CRIT: Graph-Based Automatic Data Synthesis to Enhance Cross-Modal Multi-Hop Reasoning [28.037871701844654]
  We introduce CRIT, a new dataset and benchmark built with a graph-based automatic pipeline for generating complex cross-modal reasoning tasks.<n>CRIT consists of diverse domains ranging from natural images, videos, and text-rich sources, and includes a manually verified test set for reliable evaluation.<n>Models trained on CRIT show significant gains in cross-modal multi-hop reasoning, including strong improvements on SPIQA and other standard multimodal benchmarks.
  arXiv  Detail & Related papers  (2026-04-02T05:23:18Z)
• Reasoning Theater: Disentangling Model Beliefs from Chain-of-Thought [11.955186033088351]
  We provide evidence of performative chain-of-thought (CoT) in reasoning models.<n>We compare activation probing, early forced answering, and a CoT monitor across two large models.<n>We contrast this with genuine reasoning in difficult multihop GPQA-Diamond questions.
  arXiv  Detail & Related papers  (2026-03-05T18:55:16Z)
• Less is More Tokens: Efficient Math Reasoning via Difficulty-Aware Chain-of-Thought Distillation [82.2288581878096]
  We present a framework for difficulty-aware reasoning that teaches models to dynamically adjust reasoning depth based on problem complexity.<n>We show that models can be endowed with such dynamic inference pathways without any architectural modifications.
  arXiv  Detail & Related papers  (2025-09-05T16:40:13Z)
• DRQA: Dynamic Reasoning Quota Allocation for Controlling Overthinking in Reasoning Large Language Models [28.90035967715762]
  Reasoning large language models (RLLMs) have recently demonstrated remarkable capabilities by performing structured and multi-step reasoning.<n>We propose Dynamic Reasoning Quota Allocation (DRQA), a novel method that transfers the benefits of resource competition from batch processing to single-question inference.
  arXiv  Detail & Related papers  (2025-08-25T08:47:36Z)
• The Challenge of Teaching Reasoning to LLMs Without RL or Distillation [31.973226821366325]
  Reasoning-capable language models achieve state-of-the-art performance in diverse complex tasks by generating long, explicit Chain-of-Thought traces.<n>We ask whether long CoT can be induced in a base model using only prompting or minimal tuning.<n>The resulting model outperforms the much larger textttQwen2.5-Math-72B-Instruct, showing that a handful of high-quality examples can unlock strong reasoning capabilities.
  arXiv  Detail & Related papers  (2025-07-14T01:14:50Z)
• ConciseHint: Boosting Efficient Reasoning via Continuous Concise Hints during Generation [74.37307916314407]
  We propose a framework dubbed ConciseHint, which continuously encourages the reasoning model to speak concisely.<n>Experiments on the state-of-the-art LRMs, including DeepSeek-R1 and Qwen-3 series, demonstrate that our method can effectively produce concise reasoning.
  arXiv  Detail & Related papers  (2025-06-23T16:20:44Z)
• Fractional Reasoning via Latent Steering Vectors Improves Inference Time Compute [60.151643048803145]
  We propose Fractional Reasoning, a framework that enables continuous control over reasoning intensity at inference time.<n>Our method operates by extracting the latent steering vector associated with deeper reasoning and reapplying it with a tunable scaling factor.<n> Experiments on GSM8K, MATH500, and GPQA demonstrate that Fractional Reasoning consistently improves performance across diverse reasoning tasks and models.
  arXiv  Detail & Related papers  (2025-06-18T21:15:59Z)
• When Reasoning Meets Compression: Understanding the Effects of LLMs Compression on Large Reasoning Models [18.37440002326957]
  We investigate how the reasoning capabilities of large reasoning models are compromised during compression.<n>We benchmark quantized, distilled, and pruned DeepSeek-R1 models on four reasoning datasets.<n>Overall, we find dynamically quantized 2.51-bit R1 reaches close-to-R1 performance.
  arXiv  Detail & Related papers  (2025-04-02T05:17:46Z)
• Self-Training Elicits Concise Reasoning in Large Language Models [23.475414693530965]
  Chain-of-thought (CoT) reasoning has enabled large language models (LLMs) to utilize additional computation through intermediate tokens.<n>We propose simple fine-tuning methods which leverage self-generated concise reasoning paths.<n>Our method achieves a 30% reduction in output tokens, across five model families on GSM8K and MATH, while maintaining average accuracy.
  arXiv  Detail & Related papers  (2025-02-27T14:14:50Z)
• STEP: Enhancing Video-LLMs' Compositional Reasoning by Spatio-Temporal Graph-guided Self-Training [87.58996020705258]
  Video Large Language Models (Video-LLMs) have recently shown strong derivation in basic video understanding tasks.<n>Video-LLMs struggle with compositional reasoning that requires multi-step explicit-temporal inference across object relations, interactions and events.<n>We propose STEP, a novel graph-guided self-training method that enables VideoLLMs to generate reasoning-rich finetuning data from any raw videos to improve itself.
  arXiv  Detail & Related papers  (2024-11-29T11:54:55Z)
• Selection-p: Self-Supervised Task-Agnostic Prompt Compression for Faithfulness and Transferability [67.77534983324229]
  In this paper, we investigate the ability of Large Language Models to develop a unified compression method that discretizes uninformative tokens. Experiments show Selection-p achieves state-of-the-art performance across numerous classification tasks. It exhibits superior transferability to different models compared to prior work.
  arXiv  Detail & Related papers  (2024-10-15T17:05:25Z)
• Perception Compressor: A Training-Free Prompt Compression Framework in Long Context Scenarios [17.720102137585503]
  Perception is a training-free prompt compression framework for large language models.<n>It includes a perception retriever that leverages guiding questions and instruction to retrieve the most relevant demonstrations.<n>We conduct extensive experiments on long context, benchmarks, iSie, LongBench, and MuSiQue.
  arXiv  Detail & Related papers  (2024-09-28T07:13:33Z)
• Teaching Broad Reasoning Skills via Decomposition-Guided Contexts [50.114651561111245]
  Question-answering datasets require a broad set of reasoning skills. We show how to use question decompositions to teach these broad reasoning skills in a robust fashion.
  arXiv  Detail & Related papers  (2022-05-25T05:13:21Z)
• Maieutic Prompting: Logically Consistent Reasoning with Recursive Explanations [71.2950434944196]
  We develop Maieutic Prompting, which infers a correct answer to a question even from the noisy and inconsistent generations of language models. Maieutic Prompting achieves up to 20% better accuracy than state-of-the-art prompting methods.
  arXiv  Detail & Related papers  (2022-05-24T06:36:42Z)
• ReasonBERT: Pre-trained to Reason with Distant Supervision [17.962648165675684]
  We present ReasonBert, a pre-training method that augments language models with the ability to reason over long-range relations and multiple, possibly hybrid contexts. Different types of reasoning are simulated, including intersecting multiple pieces of evidence, bridging from one piece of evidence to another, and detecting unanswerable cases.
  arXiv  Detail & Related papers  (2021-09-10T14:49:44Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.