Benchmarking Language Model Performance on 5th Gen Xeon at GCP
We’re on a journey to advance and democratize artificial intelligence through open source and open science.
More from the blog
Transfer learning for genomic prediction in underrepresented populations
General Science
A connectomics milestone: Mapping the complete male fruit fly brain
General Science
Introducing WeatherNext 3, our most advanced and accurate global weather AI model
Training a coding model to paint watercolours with TRL and OpenEnv
Related papers
Experimenting with Multi-Agent Software Development: Towards a Unified Platform
Large language models are redefining software engineering by implementing AI-powered techniques throughout the whole software development process, including requirement gathering, software architecture, code generation, testing, and deployment. However, it is still difficult to develop a cohesive platform that consistently produces the best outcomes across all stages. The objective of this study is to develop a unified platform that utilizes multiple artificial intelligence agents to automate the process of transforming user requirements into well-organized deliverables. These deliverables include user stories, prioritization, and UML sequence diagrams, along with the modular approach to APIs, unit tests, and end-to-end tests. Additionally, the platform will organize tasks, perform security and compliance, and suggest design patterns and improvements for non-functional requirements. We allow users to control and manage each phase according to their preferences. In addition, the platform provides security and compliance checks following European standards and proposes design optimizations. We use multiple models, such as GPT-3.5, GPT-4, and Llama3 to enable to generation of modular code as per user choice. The research also highlights the limitations and future research discussions to overall improve the software development life cycle. The source code for our uniform platform is hosted on GitHub, enabling additional experimentation and supporting both research and practical uses. \end
Bridging Humans and LLMs: Investigating Human-AI Collaboration in Multi-agent Requirements Analysis for Organizational AI Adoption
Context: Organizations adopting Artificial Intelligence (AI) face challenges in eliciting and analyzing requirements that align with strategic objectives, especially when human oversight and iterative refinement are needed. Large Language Models (LLMs)-based Multi-agent systems provide a potential solution by supporting structured and collaborative Requirements Engineering (RE) processes for AI adoption planning. Objective: The objective of this study is to investigate whether a multi-agent system, built on LLMs and supported by human input, can assist in requirements analysis for AI adoption. Method: We used a mixed-method approach: (i) designed and developed a multi-agent system to support the generation and prioritization of requirements for AI adoption, (ii) conducted multiple case studies with four companies to evaluate the system, and (iii) collected data through post-session questionnaires from nine participants and follow-up interviews, one per company. Results: Questionnaire and interview findings together indicate that the system may assist in identifying relevant and goal-aligned requirements. Seven participants considered the generated requirements relevant, and six found them aligned with organizational goals. Participants noted that iterative feedback improved completeness and feasibility, often within two feedback rounds. Both data sources show that human input was essential to clarify technical details, ensure contextual accuracy, and validate prioritization results. Participants from all companies also identified usability, transparency, and scalability as areas requiring further refinement for broader organizational use. Conclusions: LLM-based multi-agent systems can support strategic AI planning by enabling iterative refinement with human experts. Future work will include more interviews with stakeholders and adjustments to system features to improve transparency, usability, and scalability.
Constitutional Midtraining: Content Presence Drives Alignment Gains
Post-training alignment is often shallow, eroding under fine-tuning. It remains untested as to whether constitutional midtraining interventions can produce durable alignment when cleanly isolated from post-training. We build a 394M-token constitutional corpus from Anthropic's Constitution and apply constitutional midtraining at 120B scale, where principled, values-based content is inserted into midtraining. A 2x2 design (curriculum ordering x deliberative reasoning) was used to produce four constitutionally midtrained conditions, plus a control, which were evaluated on self-generated and established benchmarks including alignment under pressure, value conflict resolution, blackmail, and emergent misalignment. All models were evaluated across three stages: post-midtraining, post-SFT, and post-benign fine-tuning. Constitutionally midtrained models outperformed the control on alignment generalization and durability, notably on blackmail: SFT instilled a blackmail propensity in all models, but constitutional midtraining blunted it, with the advantage surviving benign fine-tuning (-17.5pp). This durability did not extend to settings that required active resistance to in-context pressure or conflict, where the advantage attenuates after SFT. The presence of constitutional content at midtraining also mattered more than its structure, and constitutional midtraining incurred no capability cost, on average, at any stage (MMLU, ARC-Easy, piqa, GSM8K). A modest amount of constitutional content at midtraining could therefore yield broad, persistent alignment gains, offering a cheap, complementary addition to SFT-centered pipelines. Code, data, and models are available.
“Failures” to be celebrated: an analysis of major pivots of software startups
In the context of software startups, project failure is embraced actively and considered crucial to obtain validated learning that can lead to pivots. A pivot is the strategic change of a business concept, product or the different elements of a business model. A better understanding is needed on different types of pivots and different factors that lead to failures and trigger pivots, for software entrepreneurial teams to make better decisions under chaotic and unpredictable environment. Due to the nascent nature of the topic, the existing research and knowledge on the pivots of software startups are very limited. In this study, we aimed at identifying the major types of pivots that software startups make during their startup processes, and highlighting the factors that fail software projects and trigger pivots. To achieve this, we conducted a case survey study based on the secondary data of the major pivots happened in 49 software startups. 10 pivot types and 14 triggering factors were identified. The findings show that customer need pivot is the most common among all pivot types. Together with customer segment pivot, they are common market related pivots. The major product related pivots are zoom-in and technology pivots. Several new pivot types were identified, including market zoom-in, complete and side project pivots. Our study also demonstrates that negative customer reaction and flawed business model are the most common factors that trigger pivots in software startups. Our study extends the research knowledge on software startup pivot types and pivot triggering factors. Meanwhile it provides practical knowledge to software startups, which they can utilize to guide their effective decisions on pivoting.