Hacker Papers
- 51
Alice in Wonderland: Simple Tasks Showing Complete Reasoning Breakdown in State-Of-the-Art Large Language Models
Hacker News threads
- Jun 2024Simple tasks showing reasoning breakdown in state-of-the-art LLMs
- Jun 2024AI Appears to Rapidly Be Approaching Brick Wall Where It Can't Get Smarter
- Jun 2024Novel prompting approach for Alice in Wonderland problem
- Jul 2024Reasoning in Large Language Models: A Geometric Perspective
- Sep 2024Reflection 70B, the top open-source model
- Oct 2024Understanding the Limitations of Mathematical Reasoning in LLMs
- Dec 2024Genie 2: A large-scale foundation world model
- Dec 2024Genie 2: A large-scale foundation world model
- Apr 2025Ask HN: Share your AI prompt that stumps every model
- 52
Universal adversarial perturbations
Hacker News threads
- Oct 2016Universal adversarial perturbations
- Nov 2016How close are we to solving vision?
- Nov 2016Show HN: Interactive deep convnet visualization for Keras and Tensorflow
- Jan 2017OpenNMT: Open-Source Neural Machine Translation
- Jun 2017Noninvasive Deep Brain Stimulation via Temporally Interfering Electric Fields
- Jul 2017Robust Adversarial Examples
- Oct 2017AI Model Fundamentally Cracks Captchas, Scientists Say
- Oct 2017One pixel attack for deceiving deep neural networks
- Nov 2018Adversarial Examples That Fool Both Computer Vision and Time-Limited Humans
- 53
Reasoning or Reciting? Exploring the Capabilities and Limitations of Language Models Through Counterfactual Tasks
Hacker News threads
- Sep 2023Exploring Capabilities, Limitations of Language Models by Counterfactual Tasks
- Jul 2023AI Safety and the Age of Dislightenment
- Dec 2023Will scaling work?
- Jun 2024The illusion of state in state-space models
- Jul 2024Tabloid: A clickbait headline programming language (2021)
- Aug 2024Markov chains are funnier than LLMs
- Aug 2024Markov chains are funnier than LLMs
- Nov 2024DeepThought-8B: A small, capable reasoning model
- 54
The Geometry of Truth: Emergent Linear Structure in Large Language Model Representations of True/False Datasets
Hacker News threads
- Oct 2023The Geometry of Truth: Emergent Linear Structure in How LLMs Represent Truth
- Oct 2023"Hallucinating" AIs Sound Creative, but Let's Not Celebrate Being Wrong
- Nov 2023Ask HN: What is your standard for judging when AGI exists?
- Nov 2023Rats have an imagination, new research suggests
- Nov 2023HN thread 38355930
- May 2024Google scrambles to manually remove weird AI answers in search
- Sep 2024AI is growing faster than companies can secure it, warn industry leaders
- Sep 2024GPTs and Hallucination
- Aug 2025Persona vectors: Monitoring and controlling character traits in language models
- 55
Practical Black-Box Attacks against Machine Learning
Hacker News threads
- Feb 2016Practical Attacks against Deep Learning Systems using Adversarial Examples
- Jun 2020Practical Black-Box Attacks Against Machine Learning
- Aug 2017Practical Black-Box Attacks Against Machine Learning
- Apr 2016I’m not a human: Breaking the Google reCAPTCHA [pdf]
- Aug 2017Why Deep Learning surprises me
- Nov 2017How Adversarial Attacks Work
- Jan 2018Audio Adversarial Examples
- Mar 2018AI Has a Hallucination Problem That's Proving Tough to Fix
- Oct 2019Introduction to Adversarial Machine Learning
- 56
Let Me Speak Freely? A Study on the Impact of Format Restrictions on Performance of Large Language Models
Hacker News threads
- Aug 2024A Study on the Impact of Structured Output on Performance of LLMs
- Aug 2024A Study on the Impact of Format Restrictions on Performance of LLM
- Aug 2024A Study on the Impact of Format Restrictions on Performance of LLMs
- Aug 2024Structured Outputs in the API
- Sep 2024Show HN: Best Practices for Using Structured Output from LLMs
- Sep 2025Sampling and structured outputs in LLMs
- Jul 2026How to stop Claude from saying load-bearing
- Aug 2026Why does Opus 5 feel worse to work with?
- Oct 2026Fixing Tool-Call Accuracy in vLLM
- 57
Language Models Use Trigonometry to Do Addition
Hacker News threads
- Feb 2025Language Models Use Trigonometry to Do Addition
- Feb 2025Pre-Trained Large Language Models Use Fourier Features for Addition (2024)
- Apr 2025AI coding and the peanut butter and jelly problem
- Oct 2025Reasoning is not model improvement
- Aug 2026Models Are Getting Dumber on Purpose
- Sep 2026US Military had close call after using AI for hallucinated intelligence report
- Oct 2026How accurately calibrated is Jev?
- 58
Questioning Representational Optimism in Deep Learning: The Fractured Entangled Representation Hypothesis
Hacker News threads
- May 2025Questioning Representational Optimism in Deep Learning
- Jun 2025Questioning Representational Optimism in Deep Learning
- Nov 2025Questioning Representational Optimism in Deep Learning
- May 2025Questioning Representational Optimism in Deep Learning
- Jun 2025Vision Language Models Are Biased
- Nov 2025The space of minds
- Feb 2026An autopsy of AI-generated 3D slop
- Aug 2026Small Models Have Arrived
- Sep 2026A beginning for mathematics
- 59
An Introduction to Probabilistic Programming
Hacker News threads
- Oct 2018An Introduction to Probabilistic Programming
- Oct 2021An Introduction to Probabilistic Programming
- Mar 2026An Introduction to Probabilistic Programming (2018)
- Oct 2018Google DeepMind Founder Demis Hassabis: Three Truths about AI
- Nov 2018Karl Friston: a neuroscientist who might hold the key to true AI
- Jun 2019A general-purpose probabilistic programming system with programmable inference
- Dec 2020Foundations of Probabilistic Programming
- Mar 2022Probabilistic Models of Cognition
- Aug 2023A History of Clojure (2020) [pdf]
- 60
Relating transformers to models and neural representations of the hippocampal formation
Hacker News threads
- Apr 2022Relating Transformers to Modes and Neural Representations of the Hippocampal
- May 2022Relating transformers to neural representations of hippocampal formation
- Apr 2022Relating Transformers to Neural Representations of the Hippocampal Formation
- Apr 2022Playing with DALL-E 2
- Dec 2022Ask HN: How to get back into AI?
- Jan 2024Understand how transformers work by demystifying the math behind them
- Jun 2024A most profound video game: a good cognitive aid for research
- Sep 2024Kolmogorov-Arnold networks may make neural networks more understandable
- Nov 2025The Case That A.I. Is Thinking
- 61
s1: Simple test-time scaling
Hacker News threads
- Jun 2025S1: Simple Test-Time Scaling
- Feb 2025Test-time scaling new approach: extra test-time compute improves LLM reasoning
- Feb 2025s1: Simple Test-Time Scaling
- Feb 2025Chat is a bad UI pattern for development tools
- Feb 2025S1: A $6 R1 competitor?
- Feb 2025Researchers created an open rival to OpenAI's o1 'reasoning' model for under $50
- Feb 2025LIMO: Less Is More for Reasoning
- Mar 2025QwQ-32B: Embracing the Power of Reinforcement Learning
- Apr 2025Chain of Recursive Thoughts: Make AI think harder by making it argue with itself
- Apr 2026TinyLoRA – Learning to Reason in 13 Parameters
- Jul 2026Controlling Reasoning Effort in LLMs
- 62
Mind Your Tone: Investigating How Prompt Politeness Affects LLM Accuracy (short paper)
Hacker News threads
- Oct 2025Impolite LLM prompts consistently outperform polite ones
- Oct 2025Investigating How Prompt Politeness Affects LLM Accuracy
- May 2026Prompt Politeness Affects LLM Accuracy
- May 2026The User Is Visibly Frustrated
- Jun 2026Claude Fable 5
- Jul 2026I used to love Claude, but the latest models are slowly ruining it
- Jul 2026Discovering Cryptographic Weaknesses with Claude
- Aug 2026Sol Loves to Cheat
- Aug 2026The growing divide between AI hype and software engineering reality
- Oct 2026Anthropic bans users from abusive or cruel behavior towards Claude
- Oct 2026Anthropic asks users to stop being mean to Claude
- 63
Refusal in Language Models Is Mediated by a Single Direction
Hacker News threads
- Jun 2024Refusal in language models is mediated by a single direction
- May 2026Refusal in Language Models Is Mediated by a Single Direction
- Jul 2024Explainability is not a game
- Nov 2025Heretic: Automatic censorship removal for language models
- Nov 2025Heretic: Automatic censorship removal for language models
- Jul 2026OpenAI and Hugging Face partner to address security incident
- Sep 2026US Military had close call after using AI for hallucinated intelligence report
- Sep 2026Pirate Face Rescues LLM Models from Deletion
- Sep 2026Pirate Face Rescues LLM Models from Deletion
- Oct 2026IBM CEO says open weight models are now good and cheap on-prem
- 64
Discovering Latent Knowledge in Language Models Without Supervision
Hacker News threads
- Dec 2022Discovering Latent Knowledge in Language Models Without Supervision
- Dec 2022Discovering Latent Knowledge in Language Models Without Supervision
- Dec 2022AlphaCode Attention Visualization
- Jan 2023ChatRWKV, like ChatGPT but powered by the RWKV (RNN-based, open) language model
- Feb 2023Microsoft Adds ChatGPT AI Technology to Bing Search Engine
- Feb 2023ArxivGPT: Chrome extension that summarizes arxived research papers using ChatGPT
- Feb 2023ChatGPT is falsely recommending us for a service we don't provide
- Feb 2023ChatGPT get-rich-quick schemes are coming for magazines, Amazon, and YouTube
- Jul 2024ChatGPT Isn't 'Hallucinating'–It's Bullshitting – Scientific American
- Apr 2025GPT o3 frequently fabricates actions, then elaborately justifies these actions
- 65
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
- 66
Quiet-STaR: Language Models Can Teach Themselves to Think Before Speaking
Hacker News threads
- Mar 2024Quiet-STaR: Language Models Can Teach Themselves to Think Before Speaking
- Mar 2024How Chain-of-Thought Reasoning Helps Neural Networks Compute
- Mar 2024How Chain-of-Thought Reasoning Helps Neural Networks Compute
- Mar 2024Can Demis Hassabis save Google?
- Apr 2024Mixture-of-Depths: Dynamically allocating compute in transformers
- May 2024The "Chain of Thought" Delusions
- Aug 2024OpenAI shows 'Strawberry' to feds, races to launch it
- Sep 2024Learning to Reason with LLMs
- Sep 2024g1: Using Llama-3.1 70B on Groq to create o1-like reasoning chains
- 67
Giraffe: Using Deep Reinforcement Learning to Play Chess
Hacker News threads
- Sep 2015Giraffe: Using Deep Reinforcement Learning to Play Chess
- Sep 2015Giraffe: Using Deep Reinforcement Learning to Play Chess
- Sep 2015Giraffe: Using Deep Reinforcement Learning to Play Chess
- Sep 2015Deep Learning Machine Teaches Itself International Master Level Chess in 72 HRs
- Sep 2015In just 72 hours a deep learning engine has learned chess to Master level
- Mar 2016Spawkfish: Neural network chess
- Oct 2017AlphaGo Zero: Learning from scratch
- Nov 2017Alpha Go Zero: How and Why It Works
- Nov 2017AI Uses Less Than Two Minutes of Videogame Footage to Recreate Game Engine
- Dec 2017Mastering Chess and Shogi by Self-Play with General Reinforcement Learning
- 68
Hyena Hierarchy: Towards Larger Convolutional Language Models
Hacker News threads
- Apr 2023Hyena Hierarchy: Towards Larger Convolutional Language Models
- Apr 2023Hyena a more efficient convolutional LLM when compared to GPT
- Mar 2023Hyena Hierarchy: Towards Larger Convolutional Language Models
- Mar 2023CoLT5: Faster Long-Range Transformers With Conditional Computation
- Apr 2023From Deep to Long Learning
- Apr 2023From Deep to Long Learning
- Apr 2023Stability AI Launches the First of Its StableLM Suite of Language Models
- Apr 2023New technology could blow away GPT-4 and everything like it
- May 2023OpenAI's plans according to sama
- Sep 2023The Eleuther AI Mafia
- Feb 2025The FFT Strikes Back: An Efficient Alternative to Self-Attention
- 69
Fast Inference from Transformers via Speculative Decoding
Hacker News threads
- Sep 2023Fast Inference from Transformers via Speculative Decoding
- Jul 2025Playing with Open Source LLMs
- Feb 2026Raspberry Pi Pico 2 at 873.5MHz with 3.05V Core Abuse
- Apr 2026We got 207 tok/s with Qwen3.5-27B on an RTX 3090
- May 2026Accelerating Gemma 4: faster inference with multi-token prediction drafters
- Jun 2026DeepSeek open-sources inference optimizations with 60–85% faster generation [pdf]
- Jun 2026DeepSeek open-sources inference optimizations with 60–85% faster generation [pdf]
- Jun 2026DeepSeek open-sources inference optimizations with 60–85% faster generation [pdf]
- 70
Executable Code Actions Elicit Better LLM Agents
Hacker News threads
- May 2024Executable Code Actions Elicit Better LLM Agents
- Jan 2025Show HN: Gemini 2 with unrestricted code execution
- Jul 2025Tools: Code Is All You Need
- Aug 2025Universal Tool Calling Protocol (UTCP)
- Dec 2025Anthropic acquires Bun
- Aug 2026Mistral Patent for "Code implemented tool calls"
- Aug 2026Mistral Patent for "Code implemented tool calls"
- Aug 2026Mistral Patent for "Code implemented tool calls"
- 71
xVal: A Continuous Numerical Tokenization for Scientific Language Models
Hacker News threads
- Oct 2023Single Digit Tokenization improves LLM Math abilities by up to 70x
- Oct 2023Single Digit tokenization improves LLM math abilities by up to 70x
- Nov 2023Telling GPT-4 you're scared or under pressure improves performance
- Nov 2023Sutskever: OpenAI board doing its mission to build AGI that benefits all
- Nov 2023Sam Altman's ouster was precipitated by letter to board about AI breakthrough
- Nov 2023Sam Altman's ouster was precipitated by letter to board about AI breakthrough
- Mar 2024Integer Tokenization Is Insane (2023)
- Sep 2024AGI is far from inevitable
- Oct 2024Probably pay attention to tokenizers
- 72
Deduplicating Training Data Makes Language Models Better
Hacker News threads
- May 2022Google fires another AI researcher who reportedly challenged findings
- Aug 2022Chinchilla's Wild Implications
- Aug 2022Exploring 12M of the 2.3B images used to train Stable Diffusion
- Nov 2022Ask HN: What can I personally do to fight AGAINST the Copilot class action?
- Jun 2024Large language model data pipelines and Common Crawl
- Jan 202530% drop in O1-preview accuracy when Putnam problems are slightly variated
- 73
Alignment faking in large language models
Hacker News threads
- Dec 2024AIs Will Increasingly Fake Alignment
- Feb 2025Efficient Reasoning with Hidden Thinking
- Feb 2025Scaling up test-time compute with latent reasoning: A recurrent depth approach
- Aug 2025Is chain-of-thought AI reasoning a mirage?
- Feb 2026An AI Agent Published a Hit Piece on Me – The Operator Came Forward
- Mar 2026Sarvam 105B, the first competitive Indian open source LLM
- 74
AutoCodeBench: Large Language Models are Automatic Code Benchmark Generators
Hacker News threads
- Mar 2026Go is the best language for agents
- Apr 2026Familiarity is the enemy: On why Enterprise systems have failed for 60 years
- May 2026If AI Writes Your Code, Why Use Python?
- May 2026If AI Writes Your Code, Why Use Python?
- Jul 2026How Our Rust-to-Zig Rewrite Is Going
- Jul 2026Elixir-lang.org has a new design
- Sep 2026Topcoat is pushing the boundary of server applications with Rust
- 75
RWKV: Reinventing RNNs for the Transformer Era
Hacker News threads
- May 2023RWKV: Reinventing RNNs for the Transformer Era
- Jul 2023Scaling Transformers to 1B Tokens
- Nov 2023Is anyone using self hosted LLM day to day and training it like a new employee
- Dec 2023Paving the way to efficient architectures: StripedHyena-7B
- Jan 2024Eagle 7B: Soaring past Transformers
- Jan 2024Eagle 7B: Soaring past Transformers
- Feb 2025Ask HN: Is anybody building an alternative transformer?
Data from the Hacker News archive (ODC-By), arXiv and OpenAlex. Not affiliated with Hacker News or Y Combinator.