Search: “llm”

Demystifying Numerical Instability in LLM Inference: Achieving Reproducible Inference for Mission-Critical Tasks with HEAL

arXiv:2606.21023v1 Announce Type: new Abstract: As Large Language Models (LLMs) deploy into mission-critical domains (e.g., finance, medicine, and law), output reproducibility has become a strict syst

Content type

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Research

FORGE: Fused On-Register Gradient Elimination for Memory-Efficient LLM Training

DGX agent

arXiv:2606.22932v1 Announce Type: new Abstract: Reverse-mode differentiation computes every weight gradient, writes it to memory, and only then lets the optimizer read it back. This two-phase schedule

researcharxiv-cs-lg

23 Jun 2026

Hardware

HERALD: High-Throughput Block Diffusion LLM Serving via CPU-GPU Cooperative KV Cache Retrieval

arXiv:2606.21633v1 Announce Type: new Abstract: Diffusion LLMs (dLLMs) improve GPU utilization over autoregressive decoding by generating multiple tokens per forward pass, but their KV cache still gro

hardwarearxiv-cs-lg

SegTME-UNI2: A Foundation Model-Based Framework for Generalisable Multiclass Cell Segmentation and LLM-Driven Tumour Microenvironment Characterisation in Histopathology

model-releasesarxiv-cs-cv

arXiv:2606.17702v2 Announce Type: replace Abstract: Characterising the tumour microenvironment (TME) from routine H&E-stained histology images requires simultaneous cell segmentation, feature extracti

Self-Improvement Can Self-Regress: The Rise-and-Collapse Failure Mode of LLM Self-Training

arXiv:2606.21090v1 Announce Type: cross Abstract: Self-improvement can self-regress. In REINFORCE post-training for code, a model can quickly improve on its optimized metric and then collapse within t

SingGuard: A Policy-Adaptive Multimodal LLM Guardrail with Dynamic Reasoning

model-releasesarxiv-cs-cv

arXiv:2606.22873v1 Announce Type: new Abstract: Vision-language models (VLMs) are increasingly deployed in consumer, medical, financial, and enterprise applications. This broad deployment expands the

The Metanym Game: A Self-Contained, Self-Consistent LLM Peer-Community Benchmark for Structural Intelligence

arXiv:2606.21008v1 Announce Type: cross Abstract: The metanym game is a competitive word game for LLMs that measures structural intelligence against established cognitive-science constructs. No conten

The Optimal Token Baseline: Variance Reduction for Long-Horizon LLM-RL

arXiv:2602.07078v2 Announce Type: replace Abstract: Reinforcement Learning (RL) for Large Language Models (LLMs) often suffers from training collapse in long-horizon tasks due to exploding gradient va

researcharxiv-cs-lg

UniRank: Unified Rank Allocation for Low-Rank LLM Compression

arXiv:2606.21847v1 Announce Type: new Abstract: Low-rank decomposition serves as a promising compression paradigm for large language models, however, rank allocation remains challenging: manual rules

AsFT: Anchoring Safety During LLM Fine-Tuning Within Narrow Safety Basin

arXiv:2506.08473v4 Announce Type: replace Abstract: Fine-tuning large language models (LLMs) improves performance but introduces critical safety vulnerabilities: even minimal harmful data can severely

LLMpedia: A Transparent Framework to Materialize an LLM's Encyclopedic Knowledge at Scale

arXiv:2603.24080v2 Announce Type: replace Abstract: Benchmarks like MMLU suggest flagship language models approach factuality saturation above 90%. LLMpedia shows this picture is incomplete. We materi

Schutzen: Evaluating LLM Safety in Bulgarian and German Contexts

arXiv:2606.11316v1 Announce Type: new Abstract: Large language models are increasingly deployed across professional domains, bringing hard-to-predict risks, including the generation of harmful or disr

safetyarxiv-cs-cl

Steering the Noise: Turning Random Perturbations into Effective Descent for Memory-Efficient LLM Fine-Tuning

arXiv:2601.04710v2 Announce Type: replace Abstract: Fine-tuning large language models (LLMs) achieves strong performance but is often limited by the memory overhead of backpropagation. Zeroth-order (Z

Visualizing LLM Latent Space Geometry Through Dimensionality Reduction

arXiv:2511.21594v3 Announce Type: replace Abstract: Large language models (LLMs) achieve state-of-the-art results across many natural language tasks, but their internal mechanisms remain difficult to

Tutorials

When is Your LLM Steerable?

arXiv:2606.11599v1 Announce Type: new Abstract: Activation steering offers a lightweight approach to control language models' behavior at inference time, but whether it succeeds or fails heavily depen

tutorialsarxiv-cs-cl

When Probing Accuracy Saturates, Fragility Resolves: A Complementary Metric for LLM Pre-Training Analysis

arXiv:2606.11375v1 Announce Type: cross Abstract: Standard linear probing declares a property 'encoded' when a classifier on hidden states achieves high accuracy. The protocol works well on a snapshot

3SPO: State-Score-Supervised Policy Optimization for LLM Agents

arXiv:2606.09961v1 Announce Type: cross Abstract: Training large language models (LLMs) as autonomous agents via reinforcement learning (RL) has enabled frontier models to achieve superhuman performan

AdaGC: Enhancing LLM Pretraining Stability via Adaptive Gradient Clipping

arXiv:2502.11034v3 Announce Type: replace Abstract: Loss spikes remain a persistent obstacle in large-scale language model pretraining. While previous research has attempted to identify the root cause

Continual LLM Upcycling: A Predictor-Gated Bank-Wise Sparsity Training Recipe for Dense-to-Sparse LLMs

arXiv:2606.10722v1 Announce Type: new Abstract: We study dense-to-sparse continual training as a way to construct channel-sparse large language models from dense checkpoints. Starting from a Qwen2.5-8

Divide and Cooperate: Role-Decomposed Multi-Agent LLM Training with Cross-Agent Learning Signals

arXiv:2606.10684v1 Announce Type: cross Abstract: Modern language agents which perform multi-step reasoning have shown strong performance in knowledge-intensive question answering. However, existing a

GhazalBench: Evaluating LLM Understanding and Canonical Surface-Form Access in Persian Ghazals

arXiv:2603.09979v2 Announce Type: replace Abstract: Persian poetry plays an active role in Iranian cultural practice, where verses by canonical poets such as Hafez are frequently quoted, paraphrased,

Less Context, Better Agents: Efficient Context Engineering for Long-Horizon Tool-Using LLM Agents

arXiv:2606.10209v1 Announce Type: new Abstract: Large language models deployed as autonomous agents for enterprise workflows face a key challenge: verbose tool responses from enterprise systems can ca

LLM-Based Code Documentation Generation and Multi-Judge Evaluation

arXiv:2606.09852v1 Announce Type: cross Abstract: High-quality source code documentation is vital yet often neglected, especially in critical domains like healthcare where reliability and maintainabil

Local Ai

Operator Fusion for LLM Inference on the Tensix Architecture

arXiv:2606.09879v1 Announce Type: new Abstract: This study addresses on-device inference bottlenecks of Transformer models on Tenstorrent's Tensix architecture and proposes an operator fusion strategy

local-aiarxiv-cs-lg

ProbeLLM: Automating Principled Diagnosis of LLM Failures

arXiv:2602.12966v2 Announce Type: replace Abstract: Understanding how and why large language models (LLMs) fail is becoming a central challenge as models rapidly evolve and static evaluations fall beh

Reasoning or Memorization? Direction-Aware Diversity Exploration in LLM Reinforcement Learning

arXiv:2606.10346v1 Announce Type: new Abstract: Reinforcement learning has become a key paradigm for eliciting reasoning abilities in large language models, where exploration is crucial for discoverin

ReflectiChain: Epistemic Grounding in LLM-Driven World Models for Supply Chain Resilience

arXiv:2606.10359v1 Announce Type: new Abstract: AI agents in supply chains face a fundamental epistemic gap: large language models (LLMs) interpret policies but lack physical grounding, while reinforc

Stop Early, Spend Less: Hidden-State Probes as a Practical Recipe for Streaming Moderation of LLM Outputs

arXiv:2606.10487v1 Announce Type: cross Abstract: Deploying large language models in user-facing systems requires efficient output safety filtering. Existing approaches typically rely on a separate mo

Trainable Smooth-Rotation Transforms with Learned Channel Scales for LLM Quantization

arXiv:2606.09927v1 Announce Type: cross Abstract: Post-training quantization (PTQ) is one of the most practical ways to reduce the serving cost of Large Language Models (LLMs), but activation quantiza

Two to Tango: Coupled Task-Reference Selection for Safe LLM Fine-tuning

arXiv:2606.09866v1 Announce Type: cross Abstract: Fine-tuning safety aligned large language models (LLMs) on downstream data improves adaptation but may erode learned safety behavior. Existing methods

Hardware

Unifying Data, Memory, and Compute Efficiency in LLM training: A Survey

arXiv:2606.10706v1 Announce Type: cross Abstract: Resource constraints increasingly determine what can be trained, fine-tuned, and deployed in large language models (LLMs), yet efficiency is often stu

hardwarearxiv-cs-ai

Aligned but Not Partner-Specific: Distinguishing How Multimodal LLM Agents Succeed in Reference Games Without Human-Like Conventions

arXiv:2606.08081v1 Announce Type: cross Abstract: Repeated reference games test whether interlocutors replace their initially long descriptions with shorter, partner-specific conventions grounded in s

APEX4: Efficient Pure W4A4 LLM Inference via Intra-SM Compute Rebalancing

arXiv:2606.08761v1 Announce Type: cross Abstract: W4A4 quantization promises full utilization of INT4 Tensor Cores, yet group dequantization overhead on CUDA Cores has driven existing systems to mixed

Beyond English benchmarks: clinical llm evaluation in Brazilian Portuguese

arXiv:2606.07853v1 Announce Type: cross Abstract: Large Language Models are transforming the support for clinical decision and their application in real scenarios. Yet, most benchmarks are conducted i

Brain-Prompt Injection: A Route-Safety Audit for BCI-LLM Agents

arXiv:2606.09315v1 Announce Type: cross Abstract: BCI-to-agent pipelines turn decoded neural activity into an authorization channel for tool-use agents, exposing a new attack surface we call brain-pro

CLPO: Curriculum Learning meets Policy Optimization for LLM Reasoning

arXiv:2509.25004v2 Announce Type: replace Abstract: Online reinforcement learning with verifiable rewards (RLVR) has become an effective paradigm for improving the reasoning abilities of large languag

Constrained Paraphrase Consistency for LLM Hallucination Detection

arXiv:2606.08158v1 Announce Type: cross Abstract: Large language models (LLMs) can generate factually inconsistent claims, motivating accurate and scalable hallucination detectors. Prior work largely

Agents

Contract2Tool: Learning Preconditions and Effects for Reliable Tool-Augmented LLM Agents

arXiv:2606.07904v1 Announce Type: new Abstract: Tool-augmented large language model agents increasingly rely on external APIs, but standard tool schemas describe how to call a tool, not when the tool

agentsarxiv-cs-ai

Cross-LLM Consistency in Inference: Evidence from Shared Interactions

arXiv:2606.08129v1 Announce Type: new Abstract: Large language models (LLMs) differ in architecture, training data, and optimization procedures, yet they may still develop similar internal inference p

Applications

Cutting LLM Evaluation Costs with SySRs: A Bandit Algorithm that Provably Exploits Model Similarity

arXiv:2606.07726v1 Announce Type: new Abstract: Large Language Models are typically benchmarked by evaluating every model on every test query. For practitioners seeking the best model to deploy, this

applicationsarxiv-cs-lg

EinSort: Sorting is All We Need for Tensorizing LLM

arXiv:2606.08565v1 Announce Type: cross Abstract: Tensor networks provide efficient representations for compressing large neural networks. By carefully designing shapes and topologies, they can signif

FlashCP: Load-Balanced Communication-Efficient Context Parallelism for LLM Training

arXiv:2606.08476v1 Announce Type: cross Abstract: Context parallelism (CP) is essential for training large-scale, long-context language models, as it partitions sequences to reduce memory overhead. Ho

Joint Structural Pruning and Mixed-Precision Quantization for LLM Compression

arXiv:2606.07819v1 Announce Type: new Abstract: Recently, the efficiency of Large Language Models (LLMs) deployment has become a critical concern in practical applications. While post-training quantiz

justine may kid about being bullied, but the answer sure sounds more like 2026 LLM-powered-AI slop than AGI to me.

Gary Marcus critiques an AI response about bullying, arguing that it exemplifies typical large language model output rather than demonstrating genuine artificial general intelligence (AGI). The post s

safetygary-marcus--x

LLM Inference at the Edge: Mobile, NPU, and GPU Performance Efficiency Trade-offs Under Sustained Load

arXiv:2603.23640v2 Announce Type: replace-cross Abstract: Deploying large language models on-device for always-on personal agents demands sustained inference from hardware tightly constrained in power

Applications

LLM-Orchestrated Conformance Checking in Stroke Care Without Computer-Interpretable Guidelines

arXiv:2606.09489v1 Announce Type: new Abstract: Objective: Conformance checking in healthcare seeks to assess whether patient care pathways adhere to clinical guidelines. However, its practical applic

applicationsarxiv-cs-ai

Mechanistic Data Attribution: Tracing the Training Origins of Interpretable LLM Units

arXiv:2601.21996v2 Announce Type: replace-cross Abstract: While Mechanistic Interpretability has identified interpretable circuits in LLMs, their causal origins in training data remain elusive. We int

PAEC: Position-Aware Entropy Calibration for LLM Reasoning in RLVR

arXiv:2606.08543v1 Announce Type: new Abstract: Reinforcement learning with verifiable rewards (RLVR) improves large language model reasoning but often suffers from rapid policy-entropy collapse, wher