2761
Short Message Compression Using LLMs
ts_sms works similarly to ts_zip . It uses a specific padding system compatible with arithmetic coding so that the message length does not need to be explicitly encoded.
New
The latest blog posts to take off on Hacker News. 6,331 posts.
2762
2763
Trying out QvQ – Qwen's new visual reasoning model
I thought we were done for major model releases in 2024, but apparently not: Alibaba’s Qwen team just dropped the Apache 2.0 licensed Qwen licensed (the license changed) QvQ-72B-Preview, “an …
2764
AIs Will Increasingly Fake Alignment
This post goes over the important and excellent new paper from Anthropic and Redwood Research, with Ryan Greenblatt as lead author, Alignment Faking in Large Language Models.
2765
2766
Narrative Jailbreaking for Fun and Profit
Posted on Monday 23 Dec 2024. 1,634 words, 1 link. By Matt Webb.
2767
2768
Exploring LoRA – Part 1: The Idea Behind Parameter Efficient Fine-Tuning
“The What and Why of Adapter based Parameter Efficient Fine Tuning: Understanding Its Purpose and Significance”
2769
Can AI do maths yet? Thoughts from a mathematician
So the big news this week is that o3, OpenAI’s new language model, got 25% on FrontierMath. Let’s start by explaining what this means.
2770
I sensed anxiety and frustration at NeurIPS 24
last week at NeurIPS’24, one extremely salient thing was the anxiety and frustration felt and expressed by late-year PhD students and postdocs who were confused by the job market that looks and feels so much different from what they expected perhaps when they were applying for PhD programs five or so years ago. and…
2771
Welcome to LLMflation - LLM inference cost is going down fast ⬇️
For LLM of equivalent performance, the inference cost is decreasing by 10x every year. What cost $60/million tokens in 2021 costs $.06/million tokens today.
2772
OpenAI O3 breakthrough high score on ARC-AGI-PUB
OpenAI o3 scores 75.7% on ARC-AGI public leaderboard.
2773
Building Effective "Agents"
Practical guidance from Anthropic on building reliable AI agents, from simple composable patterns to fully autonomous systems.
2774
A comparison to Waymo’s auto liability insurance claims at 25M miles
Understanding the safety impact of Automated Driving Systems (ADS) is crucial for their widespread adoption, yet robust real-world evaluation remains a critical area of development. By leveraging insurance industry third-party auto liability claims, we analyze the Waymo ADS performance across 25.3 million fully…
2775
The era of open voice assistants
Home Assistant Voice Preview Edition is the best way to get started with our open and privacy-focused voice assistant.
2776
2777
Alignment faking in large language models
A paper from Anthropic's Alignment Science team on Alignment Faking in AI large language models
2778
How we made our AI code review bot stop leaving nitpicky comments
Lessons learned from our challenges reducing LLM output: Discover everything that went wrong when trying to make our models say less.
2779
Waymo will bring autonomous vehicles to Tokyo
こんにちは(Konnichiwa) and hello, Japan! We're thrilled to announce that Waymo, in partnership with Nihon Kotsu and GO, will bring our autonomous vehicles to Tokyo for our first international road trip. There, our Driver will learn and adapt to left-hand traffic and new driving nuances associated with operating in one of…
2780
Veo 2: Our video generation model
Our leading video generation model
2781
Maximum likelihood estimation and loss functions
When I started learning about loss functions, I could always understand the intuition behind them. For example, the mean squared error (MSE) for regression seemed logical—penalizing large deviations from the ground-truth makes sense. But one thing always bothered me: I could never come up with those loss functions on…
2782
Quick takes on the recent OpenAI public incident write-up
OpenAI recently published a public writeup for an incident they had on December 11, and there are lots of good details in here! Here are some of my off-the-cuff observations: Saturation With thousa…
2783
Tenstorrent and the State of AI Hardware Startups
Semi-custom silicon is a bigger problem than Nvidia.
2784
Fast LLM Inference From Scratch (using CUDA)
This post is about building an LLM inference engine using C++ and CUDA from scratch without libraries.
2785
Computing Inside an AI
Since the launch of ChatGPT, explorations in two directions kicked into high gear.
2786
Byte Latent Transformer: Patches Scale Better Than Tokens
We introduce the Byte Latent Transformer (BLT), a new byte-level LLM architecture that, for the first time, matches tokenization-based LLM performance at...
2787
Sharing new research, models, and datasets from Meta FAIR
Meta FAIR is releasing new research artifacts that highlight our recent innovations in developing agents, robustness and safety, and architectures that facilitate machine learning.
2788
Elon Musk wanted an OpenAI for-profit
Elon Musk’s latest legal filing against OpenAI marks his fourth attempt in less than a year to reframe his claims. However, his own words and actions speak for themselves—in 2017, Elon not only wanted, but actually created, a for-profit as OpenAI’s proposed new structure.
2789
Phi-4: Microsoft's Newest Small Language Model Specializing in Complex Reasoning
Today we are introducing Phi-4, our 14B parameter state-of-the-art small language model (SLM) that excels at complex reasoning in areas such as math, in...
2790
Clio: A system for privacy-preserving insights into real-world AI use
Anthropic's new system, Clio, analyzes how people use AI while maintaining their privacy.