Machine Learning Research - The Batch | DeepLearning.AI (Page 4)

Santas in line with gifts and a ‘Photos with Santa’ sign.

Machine Learning Research

Agents Ascendant: LLMs evolve with agentic workflows, enabling autonomous reasoning and collaboration

The AI community laid the foundation for systems that can act by prompting large language models iteratively, leading to much higher performance across a range of applications.

Animation showcasing 7 key NLP topics visually expanding on the screen.

Machine Learning Research

When LLMs Propose Research Ideas: Stanford study finds AI matches human experts at writing research proposals

How do agents based on large language models compare to human experts when it comes to proposing machine learning research? Pretty well, according to one study.

Performance comparison for Gemini models across benchmarks.

Machine Learning Research

Multimodal Modeling on the Double: Google introduces Gemini 2.0 Flash, a faster, more capable AI model

Google’s Gemini 2.0 Flash, the first member of its updated Gemini family of large multimodal models, combines speed with performance that exceeds that of its earlier flagship model, Gemini 1.5 Pro, on several measures.

A GIF with scenes of a man at a café, a working robot, a ghost in a mirror, and a speeding truck.

Machine Learning Research

Open Video Gen Closes the Gap: Tencent releases HunyuanVideo, an open source model rivaling commercial video generators

The gap is narrowing between closed and open models for video generation.

Benchmark results for Phi-4, GPT, LLaMA-3.3, and Qwen 2.5 models.

Machine Learning Research

Phi-4 Beats Models Five Times Its Size: Microsoft’s Phi-4 blends synthetic and organic data to surpass larger models in math and reasoning benchmarks

Microsoft updated its smallest model family with a single, surprisingly high-performance model.

Graph showing how training loss affects token prediction accuracy and hallucination elimination.

Machine Learning Research

Getting the Facts Right: A memory method that reduces hallucinations in LLMs

Large language models that remember more hallucinate less.

Game character climbing a ladder with visible controls (QWASD) and health bars.

Machine Learning Research

Game Worlds on Tap: Genie 2 brings interactive 3D worlds to life

A new model improves on recent progress in generating interactive virtual worlds from still images.

o1 Family Benchmarks comparing pass rates across AIME, Codeforces, and GPQA.

Machine Learning Research

Higher Reasoning: OpenAI debuts o1 and pro mode for $200/month

OpenAI launched not only its highly anticipated o1 model but also an operating mode that enables the model to deliver higher performance — at a hefty price.

Table comparing HarmBench and AdvBench ASR performance across models and benchmarks.

Machine Learning Research

Breaking Jailbreaks: New E-DPO method strengthens defenses against jailbreak prompts

Jailbreak prompts can prod a large language model (LLM) to overstep built-in boundaries, leading it to do things like respond to queries it was trained to refuse to answer. Researchers devised a way to further boost the probability that LLMs will respond in ways that respect such limits.

Table comparing model performance on Mathvista, MMMU, ChartQA, DocVQA, and other tasks.

Machine Learning Research

Mistral’s Vision-Language Contender: Mistral unveils Pixtral Large, a rival to top vision-language models

Mistral AI unveiled Pixtral Large, which rivals top models at processing combinations of text and images.

Grounding DINO animation depicting object detection with bounding boxes on images.

Hardware

Object Detection for Small Devices: Grounding DINO 1.5, an edge device model built for faster, smarter object detection

An open source model is designed to perform sophisticated object detection on edge devices like phones, cars, medical equipment, and smart doorbells.

Bar charts comparing performance of AI models across six tasks.

Machine Learning Research

Reasoning Revealed: DeepSeek-R1, a transparent challenger to OpenAI o1

An up-and-coming Hangzhou AI lab unveiled a model that implements run-time reasoning similar to OpenAI o1 and delivers competitive performance. Unlike o1, it displays its reasoning steps.