Search: “reasoning”

arXiv:2604.15390v1 Announce Type: cross Abstract: Code deobfuscation is the task of recovering a readable version of a program while preserving its original behavior. In practice, this often requires

tutorialsarxiv-cs-ai

Content type

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Model Releases

ChatENV: An Interactive Vision-Language Model for Sensor-Guided Environmental Monitoring and Scenario Simulation

DGX agent

arXiv:2508.10635v3 Announce Type: replace Abstract: Understanding environmental changes from remote sensing imagery is vital for climate resilience, urban planning, and ecosystem monitoring. Yet, curr

model-releasesarxiv-cs-cv

20 Apr 2026

Learning to Look before Learning to Like: Incorporating Human Visual Cognition into Aesthetic Quality Assessment

arXiv:2604.15853v1 Announce Type: new Abstract: Automated Aesthetic Quality Assessment (AQA) treats images primarily as static pixel vectors, aligning predictions with human-rating scores largely thro

safetyarxiv-cs-cv

Majority Voting for Code Generation

arXiv:2604.15618v1 Announce Type: new Abstract: We investigate Functional Majority Voting (FMV), a method based on functional consensus for code generation with Large Language Models, which identifies

researcharxiv-cs-lg

Agents

MARCH: Multi-Agent Radiology Clinical Hierarchy for CT Report Generation

arXiv:2604.16175v1 Announce Type: new Abstract: Automated 3D radiology report generation often suffers from clinical hallucinations and a lack of the iterative verification found in human practice. Wh

agentsarxiv-cs-ai

MEDLEY-BENCH: Scale Buys Evaluation but Not Control in AI Metacognition

arXiv:2604.16009v1 Announce Type: new Abstract: Metacognition, the ability to monitor and regulate one's own reasoning, remains under-evaluated in AI benchmarking. We introduce MEDLEY-BENCH, a benchma

Neurosymbolic Repo-level Code Localization

arXiv:2604.16021v1 Announce Type: cross Abstract: Code localization is a cornerstone of autonomous software engineering. Recent advancements have achieved impressive performance on real-world issue be

Protecting Language Models Against Unauthorized Distillation through Trace Rewriting

arXiv:2602.15143v2 Announce Type: replace Abstract: Knowledge distillation is a widely adopted technique for transferring capabilities from LLMs to smaller, more efficient student models. However, una

researcharxiv-cs-ai

Self-Distillation as a Performance Recovery Mechanism for LLMs: Counteracting Compression and Catastrophic Forgetting

arXiv:2604.15794v1 Announce Type: cross Abstract: Large Language Models (LLMs) have achieved remarkable success, underpinning diverse AI applications. However, they often suffer from performance degra

safetyarxiv-cs-ai

SIMMER: Cross-Modal Food Image--Recipe Retrieval via MLLM-Based Embedding

arXiv:2604.15628v1 Announce Type: cross Abstract: Cross-modal retrieval between food images and recipe texts is an important task with applications in nutritional management, dietary logging, and cook

The Relic Condition: When Published Scholarship Becomes Material for Its Own Replacement

arXiv:2604.16116v1 Announce Type: cross Abstract: We extracted the scholarly reasoning systems of two internationally prominent humanities and social science scholars from their published corpora alon

good to know that a trillion dollars’ investment scaling has thoroughly solved the challenges of common sense.

Gary Marcus critiques the assumption that massive financial investment in AI scaling alone can solve fundamental challenges related to common sense reasoning in artificial intelligence systems. The po

safetygary-marcus--x

19 Apr 2026

Bayesian-LoRA: Probabilistic Low-Rank Adaptation of Large Language Models

arXiv:2601.21003v2 Announce Type: replace Abstract: Large Language Models usually put more emphasis on accuracy and therefore, will guess even when not certain about the prediction, which is especiall

researcharxiv-cs-ai

Agents

Beyond Literal Mapping: Benchmarking and Improving Non-Literal Translation Evaluation

arXiv:2601.07338v2 Announce Type: replace Abstract: Large Language Models (LLMs) have significantly advanced Machine Translation (MT), applying them to linguistically complex domains-such as Social Ne

agentsarxiv-cs-cl

CoopEval: Benchmarking Cooperation-Sustaining Mechanisms and LLM Agents in Social Dilemmas

arXiv:2604.15267v1 Announce Type: cross Abstract: It is increasingly important that LLM agents interact effectively and safely with other goal-pursuing agents, yet, recent works report the opposite tr

Create Expert Content: Deploying a Multi-Agent System with Terraform and Cloud Run

model-releasesgoogle-cloud-ai

In support of our mission to accelerate the developer journey on Google Cloud, we built Dev Signal: a multi-agent system designed to transform raw community signals into reliable technical guidance by

Graph-Based Alternatives to LLMs for Human Simulation

arXiv:2511.02135v2 Announce Type: replace Abstract: Large language models (LLMs) have become a popular approach for simulating human behaviors, yet it remains unclear if LLMs are necessary for all sim

MADE: A Living Benchmark for Multi-Label Text Classification with Uncertainty Quantification of Medical Device Adverse Events

arXiv:2604.15203v1 Announce Type: new Abstract: Machine learning in high-stakes domains such as healthcare requires not only strong predictive performance but also reliable uncertainty quantification

MARS^2: Scaling Multi-Agent Tree Search via Reinforcement Learning for Code Generation

arXiv:2604.14564v1 Announce Type: cross Abstract: Reinforcement learning (RL) paradigms have demonstrated strong performance on reasoning-intensive tasks such as code generation. However, limited traj

Psychological Steering of Large Language Models

arXiv:2604.14463v1 Announce Type: new Abstract: Large language models (LLMs) emulate a consistent human-like behavior that can be shaped through activation-level interventions. This paradigm is conver

Local Ai

Rethinking AI Hardware: A Three-Layer Cognitive Architecture for Autonomous Agents

arXiv:2604.13757v1 Announce Type: new Abstract: The next generation of autonomous AI systems will be constrained not only by model capability, but by how intelligence is structured across heterogeneou

local-aiarxiv-cs-ai

Style Amnesia: Investigating Speaking Style Degradation and Mitigation in Multi-Turn Spoken Language Models

arXiv:2512.23578v3 Announce Type: replace Abstract: In this paper, we show that when spoken language models (SLMs) are instructed to speak in a specific speaking style at the beginning of a multi-turn

Theory of Mind in Action: The Instruction Inference Task in Dynamic Human-Agent Collaboration

arXiv:2507.02935v2 Announce Type: replace Abstract: Successful human-agent teaming relies on an agent being able to understand instructions given by a (human) principal. In many cases, an instruction

Local Ai

Tight Sample Complexity Bounds for Best-Arm Identification Under Bounded Systematic Bias

arXiv:2604.14345v1 Announce Type: new Abstract: As search depth increases in autonomous reasoning and embodied planning, the candidate action space expands exponentially, heavily taxing computational

local-aiarxiv-cs-lg

Applications

Automatically Inferring Teachers' Geometric Content Knowledge: A Skills Based Approach

arXiv:2604.13666v1 Announce Type: cross Abstract: Assessing teachers' geometric content knowledge is essential for geometry instructional quality and student learning, but difficult to scale. The Van

applicationsarxiv-cs-lg

C2: Scalable Rubric-Augmented Reward Modeling from Binary Preferences

arXiv:2604.13618v1 Announce Type: new Abstract: Rubric-augmented verification guides reward models with explicit evaluation criteria, yielding more reliable judgments than single-model verification. H

Design Conditions for Intra-Group Learning of Sequence-Level Rewards: Token Gradient Cancellation

arXiv:2604.13088v1 Announce Type: new Abstract: In sparse termination rewards, intra-group comparisons have become the dominant paradigm for fine-tuning reasoning models via reinforcement learning. Ho

researcharxiv-cs-lg

DiPO: Disentangled Perplexity Policy Optimization for Fine-grained Exploration-Exploitation Trade-Off

arXiv:2604.13902v1 Announce Type: new Abstract: Reinforcement Learning with Verifiable Rewards (RLVR) has catalyzed significant advances in the reasoning capabilities of Large Language Models (LLMs).

safetyarxiv-cs-lg

Free Lunch for Unified Multimodal Models: Enhancing Generation via Reflective Rectification with Inherent Understanding

arXiv:2604.13540v1 Announce Type: new Abstract: Unified Multimodal Models (UMMs) aim to integrate visual understanding and generation within a single structure. However, these models exhibit a notable

researcharxiv-cs-cv

From Where Words Come: Efficient Regularization of Code Tokenizers Through Source Attribution

arXiv:2604.14053v1 Announce Type: new Abstract: Efficiency and safety of Large Language Models (LLMs), among other factors, rely on the quality of tokenization. A good tokenizer not only improves infe

Google’s DeepMind just released new 4B and 27B MedGemma models!

Google DeepMind released MedGemma, a collection of medical vision-language foundation models based on Gemma 3 in 4B and 27B parameter sizes, demonstrating advanced medical understanding and reasoning

model-releasesr-ollama

Indexing Multimodal Language Models for Large-scale Image Retrieval

arXiv:2604.13268v1 Announce Type: cross Abstract: Multimodal Large Language Models (MLLMs) have demonstrated strong cross-modal reasoning capabilities, yet their potential for vision-only tasks remain

InfiniteScienceGym: An Unbounded, Procedurally-Generated Benchmark for Scientific Analysis

arXiv:2604.13201v1 Announce Type: new Abstract: Large language models are emerging as scientific assistants, but evaluating their ability to reason from empirical data remains challenging. Benchmarks

Leveraging LLM-GNN Integration for Open-World Question Answering over Knowledge Graphs

arXiv:2604.13979v1 Announce Type: new Abstract: Open-world Question Answering (OW-QA) over knowledge graphs (KGs) aims to answer questions over incomplete or evolving KGs. Traditional KGQA assumes a c

Scaling Test-Time Compute to Achieve IOI Gold Medal with Open-Weight Models

arXiv:2510.14232v2 Announce Type: replace-cross Abstract: Competitive programming has become a rigorous benchmark for evaluating the reasoning and problem-solving capabilities of large language models

Why dynamically routing multi-timescale advantages in PPO causes policy collapse (and a simple decoupled fix) [R]

This Reddit post discusses a known instability in PPO when advantage estimates operating across different temporal scales (e.g., short-horizon and long-horizon returns) are dynamically routed or mixed

safetyr-machinelearning

Agents

A Multi-Agent Feedback System for Detecting and Describing News Events in Satellite Imagery

arXiv:2604.12772v1 Announce Type: new Abstract: Changes in satellite imagery often occur over multiple time steps. Despite the emergence of bi-temporal change captioning datasets, there is a lack of m

agentsarxiv-cs-cv

Advancing Multi-Agent RAG Systems with Minimalist Reinforcement Learning

arXiv:2505.17086v4 Announce Type: replace Abstract: Large Language Models (LLMs) equipped with modern Retrieval-Augmented Generation (RAG) systems often employ multi-turn interaction pipelines to inte

AGSC: Adaptive Granularity and Semantic Clustering for Uncertainty Quantification in Long-text Generation

arXiv:2604.06812v2 Announce Type: replace Abstract: Large Language Models (LLMs) have demonstrated impressive capabilities in long-form generation, yet their application is hindered by the hallucinati

Any smart human giving it real effort should score >90% on ARC-AGI-3

researchfrancois-chollet--x

ARC-AGI-3 is a new benchmark iteration designed so that any intelligent human applying genuine effort should achieve greater than 90% accuracy, maintaining the benchmark's core principle that tasks mu

Benchmarking Deflection and Hallucination in Large Vision-Language Models

arXiv:2604.12033v1 Announce Type: cross Abstract: Large Vision-Language Models (LVLMs) increasingly rely on retrieval to answer knowledge-intensive multimodal questions. Existing benchmarks overlook c

Benchmarking Foundation Models with Retrieval-Augmented Generation in Olympic-Level Physics Problem Solving

arXiv:2510.00919v3 Announce Type: replace-cross Abstract: Retrieval-augmented generation (RAG) with foundation models has achieved strong performance across diverse tasks, but their capacity for exper

Brain-DiT: A Universal Multi-state fMRI Foundation Model with Metadata-Conditioned Pretraining

arXiv:2604.12683v1 Announce Type: new Abstract: Current fMRI foundation models primarily rely on a limited range of brain states and mismatched pretraining tasks, restricting their ability to learn ge

safetyarxiv-cs-cv

Disposition Distillation at Small Scale: A Three-Arc Negative Result

arXiv:2604.11867v1 Announce Type: cross Abstract: We set out to train behavioral dispositions (self-verification, uncertainty acknowledgment, feedback integration) into small language models (0.6B to

Drawing on Memory: Dual-Trace Encoding Improves Cross-Session Recall in LLM Agents

arXiv:2604.12948v1 Announce Type: new Abstract: LLM agents with persistent memory store information as flat factual records, providing little context for temporal reasoning, change tracking, or cross-

GRACE: A Dynamic Coreset Selection Framework for Large Language Model Optimization

arXiv:2604.11810v1 Announce Type: cross Abstract: Large Language Models (LLMs) have demonstrated remarkable capabilities in natural language understanding and generation. However, their immense number

researcharxiv-cs-ai

Hard Negative Sample-Augmented DPO Post-Training for Small Language Models

model-releasesarxiv-cs-lg

arXiv:2512.19728v2 Announce Type: replace Abstract: Large language models (LLMs) continue to struggle with mathematical reasoning, and common post-training pipelines often reduce each generated soluti

Applications

Knowledge Is Not Static: Order-Aware Hypergraph RAG for Language Models

arXiv:2604.12185v1 Announce Type: new Abstract: Retrieval-augmented generation (RAG) enhances large language models by grounding outputs in retrieved knowledge. However, existing RAG methods including

applicationsarxiv-cs-cl