量子位 ·
用不着Ctrl+F了！全新AI论文工具，帮你在原文献里海里捞针
Public article from 量子位. Open the source to read the full record.
Home / Browse
Models & research
Examine models, papers and evaluations with their methods and limitations.
802 public stories
V2EX ·
[奇思妙想] 以后国内 toB 端端大模型服务会自带服务、工具吗
Public article from V2EX. Open the source to read the full record.
Hacker News ·
Can AI Agents Make Open-Ended Scientific Discovery? Evidence from Station
Hacker News recorded 2 points and 0 comments as of Oct 10, 2026 (UTC).
Hacker News ·
Show HN: Long term Memory and 50M token window for LLM
Submitted by Corbenic. Hacker News recorded 6 points and 2 comments as of Nov 5, 2026 (UTC).
OpenAI ·
Asana cuts model costs 76x in browser tests with GPT-6.1 Sol
Using GPT-6 Astra in Codex, Asana made its browser agent 76x cheaper and 5x faster in tests to offer customers more capable models.
量子位 ·
《柳叶刀》研究表明：AI 有望改善医患关系
Public article from 量子位. Open the source to read the full record.
arXiv ·
Diffu-LoRA: A Novel Low-Rank Adaptation for Personalized Diffusion Models
Public paper from arXiv. Open the source to read the full record.
arXiv ·
Wieszcz-XIX: A 3.1-Billion-Word Corpus of Pre-1918 Polish and Temporally Bounded Language Models Trained From Scratch
Public paper from arXiv. Open the source to read the full record.
arXiv ·
Large Language Model-Assisted Preparation of Transportation Management Plans: A Case Study with WisDOT WisTMP System
Public paper from arXiv. Open the source to read the full record.
arXiv ·
Cognitive Thermometers: Machine Learning and Logical Complexity
Public paper from arXiv. Open the source to read the full record.
arXiv ·
Lossy Compressive Text Autoencoders
Public paper from arXiv. Open the source to read the full record.
arXiv ·
Clarify, Then Focus: Statement Normalization for Conversation Analytics at Scale
Public paper from arXiv. Open the source to read the full record.
arXiv ·
Grammar Concept Annotation at Scale: Deployed Fine-Tuned Small Language Models Outperform Prompted Frontier Models
Public paper from arXiv. Open the source to read the full record.
arXiv ·
Real Long-Term Memory for AI: A 50-Million-Token Window That Is Faster and Cheaper Than Recompute
Public paper from arXiv. Open the source to read the full record.
arXiv ·
Disentangling Linguistic and Paralinguistic Information with Routed Sparse Autoencoders
Public paper from arXiv. Open the source to read the full record.
arXiv ·
Sparse Attention Is Matrix Approximation, Not Choosing from a Bag of Values
Public paper from arXiv. Open the source to read the full record.
arXiv ·
Stochastic Teacher Intervention for Agentic On-Policy Distillation
Public paper from arXiv. Open the source to read the full record.
arXiv ·
Large Language Models for Machine Translation Quality Annotation: Humans and Models Are Both Challenged
Public paper from arXiv. Open the source to read the full record.
arXiv ·
AI4Fire: Evaluating Large Language Models on Wildfire Tasks
Public paper from arXiv. Open the source to read the full record.
arXiv ·
When Citations Mislead? A Claim-Level Benchmark for Legal Hallucination Detection
Public paper from arXiv. Open the source to read the full record.
arXiv ·
Back in Style: A Sociolinguistic Approach to Authoring and Measuring Persona Fidelity in User Simulation
Public paper from arXiv. Open the source to read the full record.
arXiv ·
Prompts versus Rules: Auditing and Controlling Speech Naturalness Behaviors in Voice User Simulators
Public paper from arXiv. Open the source to read the full record.
arXiv ·
FedAlphaEdit: Null-Space-Aligned Merging for Collaborative Knowledge Editing
Public paper from arXiv. Open the source to read the full record.
arXiv ·
Clinician use of language models diverges from how the models are evaluated
Public paper from arXiv. Open the source to read the full record.
arXiv ·
Lapras: Latent Reasoning for Time Series Language Models
Public paper from arXiv. Open the source to read the full record.
arXiv ·
SFT-as-Context Mitigates Forgetting in Supervised Fine-Tuning
Public paper from arXiv. Open the source to read the full record.
arXiv ·
Can a System-One LLM Perform Knowledge Tracing When Few or No Learners Are Logged?
Public paper from arXiv. Open the source to read the full record.
arXiv ·
The "10th Juror": Open-Set Standpoint Screening for Bureaucratic Bias Detection
Public paper from arXiv. Open the source to read the full record.
arXiv ·
Local Prototype Reconstruction for Text-Compatible Speech-to-LLM Bridge Pretraining
Public paper from arXiv. Open the source to read the full record.
arXiv ·
LadderEdit: Edit-Level Residual Compression for Memory-Efficient Lifelong Editing of LLMs
Public paper from arXiv. Open the source to read the full record.