Search: “llm”

Ghosted Layers: Unconstrained Activation Alignment for Recovering Layer-Pruned LLMs

arXiv:2605.15491v1 Announce Type: cross Abstract: Layer pruning removes entire Transformer decoder blocks from large language models, but introduces a mismatch between the hidden state received by the

18 May 2026

Human

Content type

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Safety

Note that as of last summer Waymo said that LLMs/VLMs were experimental, and that their foundation model then could “process only a small am…

DGX agent

Note that as of last summer Waymo said that LLMs/VLMs were experimental, and that their foundation model then could “process only a small amount of image frames, does not incorporate accurate 3D sensi

safetygary-marcus--x

18 May 2026

SGR: A Stepwise Reasoning Framework for LLMs with External Subgraph Generation

arXiv:2605.16117v1 Announce Type: new Abstract: Large Language Models (LLMs) have demonstrated strong capabilities across diverse NLP applications, such as translation, text generation, and question a

18 May 2026

Toward LLMs Beyond English-Centric Development

arXiv:2605.15613v1 Announce Type: new Abstract: Through an analysis of sequences generated by open-weight large language models (LLMs), we demonstrate that LLMs are heavily biased toward English. Whil

researcharxiv-cs-cl

18 May 2026

Generative Floor Plan Design with LLMs via Reinforcement Learning with Verifiable Rewards

arXiv:2605.14117v1 Announce Type: cross Abstract: An AI system for professional floor plan design must precisely control room dimensions and areas while respecting the desired connectivity between roo

OmniDrop: Layer-wise Token Pruning for Omni-modal LLMs via Query-Guidance

arXiv:2605.14458v1 Announce Type: new Abstract: Omni-modal large language models have demonstrated remarkable potential in holistic multimodal understanding; however, the token explosion caused by hig

Prompting Policies for Multi-step Reasoning and Tool-Use in Black-box LLMs with Iterative Distillation of Experience

arXiv:2605.14443v1 Announce Type: new Abstract: The shift toward interacting with frozen, 'black-box' Large Language Models (LLMs) has transformed prompt engineering from a heuristic exercise into a c

Quantifying and Mitigating Premature Closure in Frontier LLMs

arXiv:2605.15000v1 Announce Type: cross Abstract: Premature closure, or committing to a conclusion before sufficient information is available, is a recognized contributor to diagnostic error but remai

1/5 Caching carries a deterministic assumption baked in: same input, same output. That breaks down with LLMs, and especially with agents run…

model-releasesai21-labs--x

This post discusses how traditional caching mechanisms assume deterministic behavior (identical inputs producing identical outputs), an assumption that breaks down with large language models and espec

Asynchronous Reasoning: Training-Free Interactive Thinking LLMs

arXiv:2512.10931v3 Announce Type: replace Abstract: Many state-of-the-art LLMs are trained to think before giving their answer. Reasoning can greatly improve language model capabilities, but it also m

safetyarxiv-cs-lg

Controlling Logical Collapse in LLMs via Algebraic Ontology Projection over F2

arXiv:2605.12968v1 Announce Type: cross Abstract: Do large language models internally encode ontological relations in a formally verifiable algebraic structure? We introduce Algebraic Ontology Project

Just Ask for a Table: A Thirty-Token User Prompt Defeats Sponsored Recommendations in Twelve LLMs

arXiv:2605.12772v1 Announce Type: new Abstract: Wu et al. (2026) showed that most frontier large language models (LLMs) recommend a sponsored, roughly twice-as-expensive flight when their system promp

researcharxiv-cs-cv

Local Ai

Layer-wise Representation Dynamics: An Empirical Investigation Across Embedders and Base LLMs

arXiv:2605.12714v1 Announce Type: new Abstract: Hidden states change substantially across the layers of modern language models, but most layer-wise analyses focus on one aspect of that change. We prop

local-aiarxiv-cs-lg

Sockpuppetting: Jailbreaking LLMs by Combining Prefilling with Optimization

arXiv:2601.13359v2 Announce Type: replace-cross Abstract: Prefill attacks are an effective and low-cost jailbreaking method, as they directly insert an acceptance sequence (e.g., 'Sure, here is how to

The table in HTML format for easier (and non-truncated) viewing: https://sebastianraschka.com/llm-architecture-gallery/active-parameter-rati…

model-releasessebastian-raschka--x

Sebastian Raschka shared an HTML-formatted table comparing active parameter ratios across different large language model architectures for improved readability and to avoid text truncation. The resour

Useful Memories Become Faulty When Continuously Updated by LLMs

arXiv:2605.12978v1 Announce Type: new Abstract: Learning from past experience benefits from two complementary forms of memory: episodic traces -- raw trajectories of what happened -- and consolidated

Industry

1/ Excited to report we have a new paper out @Nature today! The bottom line: training data for LLMs does not just fall from the sky - it is …

1/ Excited to report we have a new paper out @Nature today! The bottom line: training data for LLMs does not just fall from the sky - it is created in the context of existing social political institut

industryemad-mostaque--x

Beyond Parameter Aggregation: Semantic Consensus for Federated Fine-Tuning of LLMs

arXiv:2605.11857v1 Announce Type: new Abstract: Federated fine-tuning of large language models is commonly formulated as a parameter aggregation problem. However, even parameter-efficient methods requ

Evaluating the Pre-Consultation Ability of LLMs using Diagnostic Guidelines

arXiv:2601.03627v3 Announce Type: replace Abstract: We introduce EPAG, a benchmark dataset and framework designed for Evaluating the Pre-consultation Ability of LLMs using diagnostic Guidelines. LLMs

GRAFT: Graph-Tokenized LLMs for Tool Planning

arXiv:2605.11706v1 Announce Type: new Abstract: Large language models (LLMs) are increasingly used to complete complex tasks by selecting and coordinating external tools across multiple steps. This re

safetyarxiv-cs-lg

How Does Differential Privacy Affect Social Bias in LLMs? A Systematic Evaluation

arXiv:2605.11195v1 Announce Type: new Abstract: Large language models (LLMs) trained on web-scale corpora can memorize sensitive training data, posing significant privacy risks. Differential privacy (

safetyarxiv-cs-cl

Tutorials

Mind the Pause: Disfluency-Aware Objective Tuning for Multilingual Speech Correction with LLMs

arXiv:2605.12242v1 Announce Type: new Abstract: Automatic Speech Recognition (ASR) transcripts often contain disfluencies, such as fillers, repetitions, and false starts, which reduce readability and

tutorialsarxiv-cs-cl

Applications

Prompting from the bench: Large-scale pretraining is not sufficient to prepare LLMs for ordinary meaning analysis

arXiv:2510.25356v2 Announce Type: replace Abstract: In the U.S. judicial system, a widespread approach to legal interpretation entails assessing how a legal text would be understood by an `ordinary' s

applicationsarxiv-cs-cl

Reconstructing Sepsis Trajectories from Clinical Case Reports using LLMs: the Textual Time Series Corpus for Sepsis

arXiv:2504.12326v3 Announce Type: replace Abstract: Clinical case reports and discharge summaries may be the most complete and accurate summarization of patient encounters, yet they are finalized, i.e

AssayBench: An Assay-Level Virtual Cell Benchmark for LLMs and Agents

arXiv:2605.10876v1 Announce Type: cross Abstract: Recent advances in machine learning and large-scale biological data collections have revived the prospect of building a virtual cell, a computational

ConQuR: Corner Aligned Activation Quantization via Optimized Rotations for LLMs

arXiv:2605.10793v1 Announce Type: new Abstract: Large language models (LLMs) are costly to deploy due to their large memory footprint and high inference cost. Weight-activation quantization can reduce

Insight: Enhancing Mobile Accessibility for Blind and Visually Impaired Users with LLMs

arXiv:2605.09803v1 Announce Type: cross Abstract: This research paper addresses the limitations of current mobile accessibility services like TalkBack, which provide manual gesture-based sequential fe

Intrinsic Guardrails: How Semantic Geometry of Personality Interacts with Emergent Misalignment in LLMs

arXiv:2605.10633v1 Announce Type: cross Abstract: Fine-tuning Large Language Models (LLMs) on benign narrow data can sometimes induce broad harmful behaviors, a vulnerability termed emergent misalignm

'It is categorically not a victory for pure LLMs; it’s a victory for borrowing from classical AI and CS to move beyond pure LLMs.' ☝️ Co…

model-releasesgary-marcus--x

'It is categorically *not* a victory for pure LLMs; it’s a victory for borrowing from classical AI and CS to move *beyond* pure LLMs.' ☝️ Compound AI Systems alllll the way downnn 🤩🤯🤩 Claude Code (sti

LLM4Branch: Large Language Model for Discovering Efficient Branching Policies of Integer Programs

arXiv:2605.10401v1 Announce Type: new Abstract: Efficient branching policies are essential for accelerating Mixed Integer Linear Programming (MILP) solvers. Their design has long relied on hand-crafte

Hardware

LLMs for Secure Hardware Design and Related Problems: Opportunities and Challenges

arXiv:2605.10807v1 Announce Type: cross Abstract: The integration of Large Language Models (LLMs) into Electronic Design Automation (EDA) and hardware security is rapidly reshaping the semiconductor i

hardwarearxiv-cs-lg

Magis-Bench: Evaluating LLMs on Magistrate-Level Legal Tasks

arXiv:2605.08437v1 Announce Type: cross Abstract: Existing benchmarks for legal AI focus primarily on tasks where LLMs must produce legal arguments or documents, yet the capacity to judge such argumen

Applications

Position: Avoid Overstretching LLMs for every Enterprise Task

arXiv:2605.09365v1 Announce Type: new Abstract: Enterprise workloads are dominated by deterministic, structured, and knowledge-dependent tasks operating under strict cost, latency, and reliability con

applicationsarxiv-cs-ai

Re-Triggering Safeguards within LLMs for Jailbreak Detection

arXiv:2605.10611v1 Announce Type: cross Abstract: This paper proposes a jailbreaking prompt detection method for large language models (LLMs) to defend against jailbreak attacks. Although recent LLMs

Really amazing dissection. LLMs hitting rarely. Tools list is amazing.

model-releasesgary-marcus--x

Really amazing dissection. LLMs hitting rarely. Tools list is amazing. 🤩🤯🤩 Claude Code (still not AGI but biggest advance since GPT-4) is the most neurosymbolic thing I have ever seen in my life. 53 s

Scam2Prompt: A Scalable Framework for Auditing Malicious Scam Endpoints in Production LLMs

arXiv:2509.02372v3 Announce Type: replace-cross Abstract: Large Language Models have become critical to modern software development, but their reliance on uncurated web-scale datasets for training int

Soohak: A Mathematician-Curated Benchmark for Evaluating Research-level Math Capabilities of LLMs

arXiv:2605.09063v1 Announce Type: new Abstract: Following the recent achievement of gold-medal performance on the IMO by frontier LLMs, the community is searching for the next meaningful and challengi

Teaching LLMs to See Graphs: Unifying Text and Structural Reasoning

arXiv:2605.10247v1 Announce Type: new Abstract: Using Large Language Models (LLMs) to process graph-structured data is an active research area, yet current state-of-the-art approaches typically rely o

Team-Based Self-Play With Dual Adaptive Weighting for Fine-Tuning LLMs

arXiv:2605.09922v1 Announce Type: cross Abstract: While recent self-training approaches have reduced reliance on human-labeled data for aligning LLMs, they still face critical limitations: (i) sensiti

TELL-TALE: Task Efficient LLMs with Task Aware Layer Elimination

arXiv:2510.22767v3 Announce Type: replace-cross Abstract: Large Language Models (LLMs) typically come with a fixed architecture, despite growing evidence that not all layers contribute equally to ever

researcharxiv-cs-cl

The Realignment Problem: When Right becomes Wrong in LLMs

arXiv:2511.02623v2 Announce Type: replace Abstract: Post-training alignment of large language models (LLMs) relies on large-scale human annotations guided by policy specifications that change over tim

Towards Effective Theory of LLMs: A Representation Learning Approach

arXiv:2605.09294v1 Announce Type: cross Abstract: We propose Representational Effective Theory (RET), a framework for describing large language model computation in terms of learned macrostates rather

Applications

Can LLMs Take Retrieved Information with a Grain of Salt?

arXiv:2605.06919v1 Announce Type: new Abstract: Large language models have demonstrated impressive retrieval-augmented capabilities. However, a crucial area remains underexplored: their ability to app

applicationsarxiv-cs-cl

Direction-Flipped Influence Audits Reveal Hidden Structure in Moral Choices of LLMs

arXiv:2602.22831v2 Announce Type: replace-cross Abstract: Moral benchmarks for LLMs typically score models on context-free prompts, implicitly treating the measured choice rate as stable. We test this

From Standalone LLMs to Integrated Intelligence: A Survey of Compound Al Systems

arXiv:2506.04565v2 Announce Type: replace-cross Abstract: Compound AI Systems (CAIS) are an emerging paradigm that integrates large language models (LLMs) with external components, including retriever

researcharxiv-cs-cl

Important Nature Neuroscience paper shows how humans differ from LLMs. Many people currently believe that humans are just next-word predicto…

Important Nature Neuroscience paper shows how humans differ from LLMs. Many people currently believe that humans are just next-word predictors, like LLMs. But this new paper by Zou, Poeppel and Ding s

safetygary-marcus--x

LLMs are not (consistently) Bayesian: Quantifying internal (in)consistencies of LLMs' probabilistic beliefs

arXiv:2605.06915v1 Announce Type: new Abstract: Modern AI systems are being deployed in complex domains such as medicine, science, and law, where it is important that they not only produce correct ans

researcharxiv-cs-lg

Agents

LLMs Improving LLMs: Agentic Discovery for Test-Time Scaling

arXiv:2605.08083v1 Announce Type: new Abstract: Test-time scaling (TTS) has become an effective approach for improving large language model performance by allocating additional computation during infe

agentsarxiv-cs-cl