Hacker Papers
- 76
DeepSeek-V4: Towards Highly Efficient Million-Token Context Intelligence
Hacker News threads
- 77
Mixture-of-Depths: Dynamically allocating compute in transformer-based language models
Hacker News threads
- Apr 2024Mixture-of-Depths: Dynamically allocating compute in transformers
- Apr 2024Mixture-of-Depths: Dynamically allocating compute in transformer language models
- Apr 2024Mixture-of-Depths: Dynamically allocating compute in transformers
- Apr 2024Show HN: Speeding up LLM inference 2x times (possibly)
- Aug 2024Antifragility in complex dynamical systems
- Sep 2026GPT-6 Astra, looped transformers, and hidden reasoning
- 78
ReAct: Synergizing Reasoning and Acting in Language Models
Hacker News threads
- Dec 2024ReAct: Synergizing Reasoning and Acting in Language Models
- Jan 2023ReAct: Synergizing Reasoning and Acting in Language Models
- Feb 2023ReAct – interesting paper on LLM prompt engineering
- Mar 2023Visual ChatGPT
- Mar 2023TaxyAI: Open-source browser automation with GPT-4
- Apr 2023The LLama Effect: Leak Sparked a Series of Open Source Alternatives to ChatGPT
- May 2023Delimiters won’t save you from prompt injection
- Jul 2023Langchain Is Pointless
- Sep 2023ChatGPT can now see, hear, and speak – openai.com
- Jul 2025Hierarchical Reasoning Model
- Sep 2026Prompting Is Dead in 6 Months. Andrew Ng, Stanford [video]
- 79
OLMoTrace: Tracing Language Model Outputs Back to Trillions of Training Tokens
Hacker News threads
- 80
An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale
Hacker News threads
- Apr 2022An Image Is Worth 16x16 Words: Transformers for Image Recognition at Scale
- Jul 2021Deep reinforcement learning is a waste of time (2019)
- Mar 2022Transformers, originally designed to handle language, are taking on vision
- Mar 2023The Most Cited AI Papers in 2022
- Jun 2023What Is a Transformer Model?
- Feb 2024The killer app of Gemini Pro 1.5 is using video as an input
- Oct 2024Show HN: Doctly AI – Accurate AI-Powered PDF to Markdown Parser
- Jul 2025Baba Is Eval
- Sep 2026Ask HN: (Why) Was the LLM breakthrough useful for images, audio, etc.?
- 81
Position: Stop Anthropomorphizing Intermediate Tokens as Reasoning/Thinking Traces!
Hacker News threads
- May 2025(How) Do reasoning models reason?
- Apr 2025(How) Do reasoning models reason?
- Jun 2025Stop Anthropomorphizing Intermediate Tokens as Reasoning/Thinking Traces
- Oct 2025A small number of samples can poison LLMs of any size
- Nov 2025Learn Prolog Now (2006)
- Jun 2026GLM-5.2 is the new leading open weights model on Artificial Analysis
- Aug 2026Chain-of-Thought Reasoning in the Wild Is Not Always Faithful
- Sep 2026OpenAI's new reasoning technique alarms AI safety experts
- Sep 2026How concerned should we be about Astra's recurrent architecture?
- Sep 2026Formalizing Fermat's Last Theorem
- Sep 2026Caltech Mathathon – first hackathon ever devoted to research level mathematics
- 82
ExpertPrompting: Instructing Large Language Models to be Distinguished Experts
Hacker News threads
- 83
Training Large Language Models to Reason in a Continuous Latent Space
Hacker News threads
- Dec 2024Training LLMs to Reason in a Continuous Latent Space
- Apr 2026Training Large Language Models to Reason in a Continuous Latent Space [pdf]
- Dec 2024Coconut from Meta lets language models think without words
- Dec 2024Coconut by Meta AI – Better LLM Reasoning with Chain of Continuous Thought?
- Jan 2025DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via RL
- Apr 2025The Llama 4 herd
- Jun 2025Reasoning by Superposition: A Perspective on Chain of Continuous Thought
- Dec 2025Universal Reasoning Model (53.8% pass 1 ARC1 and 16.0% ARC 2)
- Jun 2026Claude Code's "extended thinking" is a summary- not authentic thinking
- Sep 2026What is Nueralese and Why is it Bad
- 84
Harnessing the Universal Geometry of Embeddings
Hacker News threads
- May 2025Harnessing the Universal Geometry of Embeddings
- Sep 2026Harnessing the Universal Geometry of Embeddings
- May 2025Text Embeddings are All Alike
- Dec 2025The universal weight subspace hypothesis
- Jan 2026Moltbook
- Mar 2026
- Jul 2026Emily Bender Sets the Record Straight on "Stochastic Parrots"
- Aug 2026Models Are Getting Dumber on Purpose
- 85
Extracting Training Data from Large Language Models
Hacker News threads
- Dec 2020Extracting Training Data from Large Language Models
- Dec 2020Extracting Training Data from Large Language Models
- Jun 2021Extracting Training Data from Large Language Models
- Dec 2020Does GPT-2 know your phone number?
- Feb 2021Google Open-Sources Trillion-Parameter AI Language Model Switch Transformer
- Jul 2021All public GitHub code was used in training Copilot
- Jun 2022Ask HN: GPT-3 reveals my full name – can I do anything?
- Apr 2023Record Labels Panic as A.I. Generated Song Becomes Hit Single
- Sep 2026“Next-token predictor” is the wrong mental model for LLMs
- 86
Tulu 3: Pushing Frontiers in Open Language Model Post-Training
Hacker News threads
- 87
Language Models are Few-Shot Learners
Hacker News threads
- May 2020GPT-3: Language Models Are Few-Shot Learners
- Jun 2024Language Models Are Few-Shot Learners
- Nov 2024Language models are few-shot learners (2020)
- Jun 2020GPT-3 and Scaling Trends
- Jul 2020Show HN: Humans vs AI – A/B testing GPT-3
- Oct 2020AI training method exceeds GPT-3 performance with fewer parameters
- Oct 2022Why Mastering Language Is So Difficult for AI
- Feb 2023Why China Didn’t Invent ChatGPT
- May 2023Michael J. Fox Foundation Announces Breakthrough for Parkinson's Biomarker
- Nov 2023Sam Altman Exposes the Charade of AI Accountability
- Sep 2026GPT-6 Astra
- 88
Chain-of-Thought Prompting Elicits Reasoning in Large Language Models
Hacker News threads
- Apr 2022Chain of Thought Prompting Elicits Reasoning in Large Language Models
- Dec 2022Chain of Thought Prompting Elicits Reasoning in Large Language Models
- Apr 2022Pathways Language Model (PaLM): Scaling to 540B parameters
- Jul 2022Am I doing it wrong? (OpenAI test)
- Mar 2023Oceania Always at War with Eastasia: Generative AI and Knowledge Pollution
- Apr 2023Teaching ChatGPT to Speak My Son’s Invented Language
- Aug 2023GPT-4 Can't Reason
- Sep 2024Training Language Models to Self-Correct via Reinforcement Learning
- Oct 2025Ortega hypothesis
- Sep 2026GPT-6 Astra
- 89
Large Language Models as Markov Chains
Hacker News threads
- Nov 2024Large Language Models as Markov Chains
- Oct 2024LLMs as Markov Chains
- Oct 2024Large Language Models as Markov Chains
- Jul 2025Problem solving using Markov chains (2007) [pdf]
- Sep 2025The wall confronting large language models
- Oct 2025BERT is just a single text diffusion step
- Nov 2025AI is a front for consolidation of resources and power
- Dec 2025Reflections on AI at the End of 2025
- May 2026Why AI Agents Cannot Change Software Systems
- Jun 2026How LLMs Work
- Sep 2026The Emergent Symbolic Structure of Artificial Neural Networks
- 90
Understanding and Mitigating Numerical Sources of Nondeterminism in LLM Inference
Hacker News threads
- Jun 2025Give Me FP32 or Give Me Death?
- Sep 2026Understanding, Mitigating Numerical Sources of Nondeterminism in LLM Inference
- Jun 2025TPU Deep Dive
- Jul 2025LLMs as Compilers
- Jul 2025Grok: Searching X for "From:Elonmusk (Israel or Palestine or Hamas or Gaza)"
- Feb 2026LLMs could be, but shouldn't be compilers
- Mar 2026Surpassing vLLM with a Generated Inference Stack
- Aug 2026What my dad taught me about AI coding in the 90s
- 91
Conditional Memory via Scalable Lookup: A New Axis of Sparsity for Large Language Models
Hacker News threads
- Mar 2026DeepSeek: Conditional Memory via Scalable Lookup
- Jan 2026Conditional Memory via Scalable Lookup
- Jan 2026RAM Coffers– I built conditional memory for LLMs 27 days before DeepSeek'sEngram
- Jan 2026Subject-based weight routing for LLMs (27 days before DeepSeek Engram)
- Mar 2026Intel Announces Arc Pro B70 and Arc Pro B65 GPUs
- Aug 2026Qwen3.8-Flash-Next: A New Architecture, Towards Ultimate Cost-Efficiency
- Sep 2026Show HN: Engram – Shared procedural runbook memory for AI coding agents via MCP
- 92
Emergent Abilities of Large Language Models
Hacker News threads
- Feb 2023Emergent Abilities of Large Language Models
- May 2024Emergent Abilities of Large Language Models
- Apr 2023Emergent Abilities of Large Language Models
- Dec 2022Ask HN: How does ChatGPT work?
- Feb 2023It Looks Like You’re Trying to Take over the World
- May 2023LLMs are not greater than the sum of their parts: Stanford researchers
- Jul 2023Llama 2 – Meta AI
- Nov 2023Telling GPT-4 you're scared or under pressure improves performance
- Mar 2024“Emergent” abilities in LLMs actually develop gradually and predictably – study
- Feb 2025Understanding Reasoning LLMs
- Aug 2026Humanity has the debate about AI consciousness backwards
- 93
Lost in the Middle: How Language Models Use Long Contexts
Hacker News threads
- Jan 2026Lost in the Middle: How Language Models Use Long Contexts (2023)
- Jul 2023Lost in the Middle: How Language Models Use Long Contexts
- Aug 2023Language models use long contexts
- Jul 2023Llama 32K Context Released by Together AI
- Aug 2023What's new in Llama 2 and how to run it locally
- Sep 2023A Comprehensive Guide for Building Rag-Based LLM Applications
- Oct 2023Microsoft says GPT 3.5 has 20B parameters?
- Nov 2023Benchmarking GPT-4 Turbo – A Cautionary Tale
- Apr 2024LLMs and the Harry Potter problem
- Jul 2026OpenAI reduces Codex Model Context Size from 372k to 272k
- Aug 2026Serve Markdown to AI Agents with Accept Headers
- 94
LLM in a flash: Efficient Large Language Model Inference with Limited Memory
Hacker News threads
- Jan 2024LLM in a Flash: Efficient Large Language Model Inference with Limited Memory
- Dec 2023LLM in a Flash: Efficient Large Language Model Inference with Limited Memory
- Dec 2023LLM in a Flash: Efficient Large Language Model Inference with Limited Memory
- Dec 2023Apple Announces LLM in a Flash: Efficient LLM Inference with Limited Memory
- Dec 2023Apple wants AI to run directly on its hardware instead of in the cloud
- Dec 2023Show HN: Emu2 – A Gemini-like open-source 37B Multimodal Model
- Dec 2023Ferret: A Multimodal Large Language Model
- Mar 2024Towards 1-bit Machine Learning Models
- Mar 2026Qwen3.5-397B at 4.74 tok/s using 5.9GB RAM
- Mar 2026iPhone 17 Pro Demonstrated Running a 400B LLM
- Aug 2026New Mac Studio with M5 Max and M5 Ultra
- 95
Large Language Models Cannot Self-Correct Reasoning Yet
Hacker News threads
- Nov 2023LLMs cannot self-correct reasoning yet
- Oct 2023LLMs can't self-correct in reasoning tasks, DeepMind study finds
- Oct 2023GPT-4 Understands
- Apr 2024The question that no LLM can answer and why it is important
- Aug 2024OpenDevin: An Open Platform for AI Software Developers as Generalist Agents
- Mar 2025Show HN: We made an MCP server so Cursor can debug Node.js on its own
- Aug 2026Dutch lawyer fined €2,300 for using AI-hallucinated citation
- 96
Inductive logic programming at 30: a new introduction
Hacker News threads
- 97
Distilling the Knowledge in a Neural Network
Hacker News threads
- Mar 2015Distilling the Knowledge in a Neural Network
- Jul 2015How Google Translate squeezes deep learning onto a phone
- Mar 2016AlphaGo beats Lee Sedol 3-0 [video]
- Apr 2017The Myth of a Superhuman AI
- Dec 2017Neural Networks in JavaScript with Deeplearn.js
- Apr 2022Scientists discover specific neurons involved in making memories
- Mar 2023Could you train a ChatGPT-beating model for $85,000 and run it in a browser?
- Mar 2026NanoGPT Slowrun: 10x Data Efficiency with Infinite Compute
- Aug 2026OpenAI: GPT 5.6 Sol price reduction (until at least Nov 21)
- 98
People who frequently use ChatGPT for writing tasks are accurate and robust detectors of AI-generated text
Hacker News threads
- Jul 2025People who frequently use ChatGPT for writing tasks can detect AI-generated text
- Jan 2025People who use ChatGPT for writing are robust detectors of AI-generated text
- May 2026People who use ChatGPT for writing are accurate detectors of AI text (2025)
- Nov 2025When your hash becomes a string: Hunting Ruby's million-to-one memory bug
- Apr 2026Ask HN: How do systems (or people) detect when a text is written by an LLM
- Apr 2026We found an undocumented bug in the Apollo 11 guidance computer code
- May 2026Americans Are Smashing Flock Cameras
- Jun 2026"No, I swear I wrote this."
- Jul 2026Ask HN: Add flag for AI-generated articles
- Aug 2026Where .env Went Wrong
- Aug 2026I'm Becoming AI-Blind
- 99
TruthfulQA: Measuring How Models Mimic Human Falsehoods
Hacker News threads
- 100
Computational Life: How Well-formed, Self-replicating Programs Emerge from Simple Interaction
Hacker News threads
- Jun 2024Computational Life: How self-replicating programs emerge from simple interaction
- Jun 2024Computational Life: How Self-Replicating Programs Emerge from Simple Interaction
- Jul 2024The data that powers AI is disappearing fast
- Jul 2024Google creates self-replicating life from digital 'primordial soup'
- Aug 2024Self-replicating programs emerge from simple interaction (in JavaScript)
- Sep 2024Brainfuck Enterprise Solutions
- Nov 2024OpenAI, Google and Anthropic are struggling to build more advanced AI
- Nov 2025Updated practice for review articles and position papers in ArXiv CS category
- Aug 2026The Computational Theory of Mind (2015)
- Aug 2026Computational Life Reactor (Based on Brainfuck)
Data from the Hacker News archive (ODC-By), arXiv and OpenAlex. Not affiliated with Hacker News or Y Combinator.