Microsoft

Models & products

Models

MAI

Trends

• Open-weight models diversify in architecture, modality, and size
• Frontier launches come under outside review and independent verification
• Agent training moves off costly live environments
• Agents improve themselves through skills, harnesses, and memory

Items

1. Oct 2

   Microsoft's ActiveSaddler

   Formulates agent-harness optimization as automated curriculum learning, modeling the evolving training curriculum as a non-stationary bandit over reusable failure-pattern arms, and lifts test Pass@1 by 4.4 points on GAIA2 and 7.5 points on Terminal-Bench 2.0 over a fixed scenario order.

2. Aug 8

   Microsoft Research's full-bandwidth transformer

   Fusing the previous top-layer hidden state with the sampled-token embedding through a gated linear unit widens the feedback channel between decoding steps, matching standard transformers trained on about 1.5x more tokens with negligible overhead and an intact KV cache.

3. Jul 24

   Microsoft Research's LLM-as-a-Coach

   An LLM judge becomes a coach whose textual feedback is distilled into the policy through on-policy context distillation, beating rubric-based RL on non-verifiable tasks while mitigating reward hacking.

4. Jul 24

   Microsoft: LLMs get lost in evolving intent

   Converting static single-turn benchmarks into multi-turn conversations with shifting intent shows strong static scores do not transfer, a blind spot for collaborative agents that static evaluation cannot see.

5. Jun 30

   Anthropic proposes a cross-lab jailbreak-severity framework

   Anthropic proposed an industry framework for scoring jailbreak severity and calibrating developer response, with Amazon, Microsoft, Google, and other Glasswing partners participating.

6. Jun 8

   Microsoft Research's Lens image model

   A 3.8B text-to-image model shows detailed training captions can substitute for raw scale, matching or beating much larger models on several benchmarks with roughly one-fifth the pre-training compute; code and checkpoints are MIT-licensed.

7. Apr 27

   Microsoft and OpenAI enter a new partnership phase

   The companies extended infrastructure commitments and go-to-market collaboration for GPT-5.x enterprise deployment, a day before OpenAI's arrival on AWS signaled a multi-cloud strategy.

8. Apr 2

   Microsoft's MAI model family

   The first models from Microsoft's MAI Superintelligence team: MAI-Transcribe-1 beats Whisper-large-v3 on all 25 FLEURS languages (3.8% average WER), MAI-Voice-1 generates a minute of speech in under a second on one GPU, and MAI-Image-2 debuted #3 on Arena.