April 2025

Data Labelling Using LLMs with Langformers

⋅

When most people think of Large Language Models (LLMs), they think of conversations, content generation, or summarization. But LLMs are also incredibly effective at data labelling — and now, with Langformers, you can easily utilize that power for you text labelling tasks. Whether you’re preparing training data, building a classifier, or just need quick annotations (maybe something around weak supervision), Langformers offers the simplest way to define labels and let LLMs do the heavy lifting. How It Works […]

Ver mais

Liked Liked
The State of Reinforcement Learning for LLM Reasoning

The State of Reinforcement Learning for LLM Reasoning

⋅

A lot has happened this month, especially with the releases of new flagship models like GPT-4.5 and Llama 4. But you might have noticed that reactions to these releases were relatively muted. Why? One reason could be that GPT-4.5 and Llama 4 remain conventional models, which means they were trained without explicit reinforcement learning for reasoning. Meanwhile, competitors such as xAI and Anthropic have added more reasoning capabilities and features into their models. For instance, both the xAI […]

Ver mais

Liked Liked

Build Your Own LLM-Powered Apps with Langformers: LLM Inference API

⋅

Today, we’re excited to showcase another powerful feature of Langformers — LLM Inference via a REST API. Whether you’re building a chatbot, automating workflows, or adding AI magic to your product, Langformers gives you everything you need to bring large language models (LLMs) into your application stack effortlessly. Let’s dive in! What is LLM Inference? LLM inference means sending a request to a language model and getting a generated response back. With Langformers, this is as easy as […]

Ver mais

Liked Liked

Chat with Open-Source LLMs Locally (Llama, Deepseek, Mistral)

⋅

Large Language Models (LLMs) are everywhere — and everyone’s using them. But there’s a big issue emerging: people are feeding sensitive emails and company data into cloud-based LLM services like ChatGPT, DeepSeek, and others. Just to draft a few emails or generate some paragraphs, confidential information is getting handed over to third-party servers. Sound a little risky? It is. Thankfully, there’s a safer alternative: open-source LLMs. Hugging Face and Ollama offer a wide range of models that you […]

Ver mais

Liked Liked
Defending against Prompt Injection with Structured Queries (StruQ) and Preference Optimization (SecAlign)

Defending against Prompt Injection with Structured Queries (StruQ) and Preference Optimization (SecAlign)

⋅

Recent advances in Large Language Models (LLMs) enable exciting LLM-integrated applications. However, as LLMs have improved, so have the attacks against them. Prompt injection attack is listed as the #1 threat by OWASP to LLM-integrated applications, where an LLM input contains a trusted prompt (instruction) and an untrusted data. The data may contain injected instructions to arbitrarily manipulate the LLM. As an example, to unfairly promote “Restaurant A”, its owner could use prompt injection to post a review […]

Ver mais

Liked Liked
Repurposing Protein Folding Models for Generation with Latent Diffusion

Repurposing Protein Folding Models for Generation with Latent Diffusion

⋅

PLAID is a multimodal generative model that simultaneously generates protein 1D sequence and 3D structure, by learning the latent space of protein folding models. The awarding of the 2024 Nobel Prize to AlphaFold2 marks an important moment of recognition for the of AI role in biology. What comes next after protein folding? In PLAID, we develop a method that learns to sample from the latent space of protein folding models to generate new proteins. It can accept compositional […]

Ver mais

Liked Liked
Simplifying Deep Temporal Difference Learning

Simplifying Deep Temporal Difference Learning

⋅

tl;dr The authors propose PQN, a simplified deep online Q-Learning that uses very small replay buffers. Normalization and parallelized sampling from vectorized environments stabilizes training without the need for huge replay buffers. PQN is competitive with more complex methods such as Rainbow, PPO-RNN, QMix while being 50x faster than traditional DQN. Introduction Temporal difference (TD) methods can be simple and efficient, but are notably unstable when combining them with neural networks or off-policy sampling. All of the following […]

Ver mais

Liked Liked

Load More

↑

Digitado © 2025