deepset Cloud, a large language model (LLM) platform, is providing insights into the precision and fidelity of responses from LLM generative AI through the Groundedness Observability Dashboard.
With the 01/2024 release, the Groundedness Observability Dashboard displays trend data for how well generative AI responses are grounded in the source documents. This feature provides a quantifiable score to measure the factuality of an LLM's output. The results serve as a guide for developers in modifying their RAG setup, fine tuning models, and altering prompts to improve accuracy and reliability of generated responses. Simplified insights into what works enables users to track how well the model can use the provided data to answer queries in a reliable manner. When tracked over time, this allows for comparisons with other widely-available LLM platforms.
deepset Cloud’s Source Reference Prediction generative response annotation is also now generally available. Response Annotation adds academic-style citations to the LLM-generated answer. Those citations reference the respective document on which a statement is based. Users can then review the source material in order to fact-check generated answers or gain a better understanding of the source data in its original context.
The combination of deepset Cloud’s Groundedness Dashboard and Source Reference Prediction gives organizations greater confidence in the quality of the responses in their LLM applications, and provides visibility when an application’s accuracy does not meet requirements.
Groundedness isn't just a useful metric for measuring the faithfulness of your LLM-generated answers to a knowledge base. It can also be used as a proxy to identify the ideal hyperparameters for your retrieval step. Optimizing the number of documents embedded in the query can reduce your LLM costs by a significant factor.
These new features emphasize deepset’s commitment to building a robust trust layer within generative AI applications. The new features effectively detect hallucinations and provide benchmarking tools, allowing users to make informed decisions about the reliability of their AI models.
These features of deepset Cloud are in General Availability.
The Latest
The use of hybrid multicloud models is forecasted to double over the next one to three years as IT decision makers are facing new pressures to modernize IT infrastructures because of drivers like AI, security, and sustainability, according to the Enterprise Cloud Index (ECI) report from Nutanix ...
Over the last 20 years Digital Employee Experience has become a necessity for companies committed to digital transformation and improving IT experiences. In fact, by 2025, more than 50% of IT organizations will use digital employee experience to prioritize and measure digital initiative success ...
While most companies are now deploying cloud-based technologies, the 2024 Secure Cloud Networking Field Report from Aviatrix found that there is a silent struggle to maximize value from those investments. Many of the challenges organizations have faced over the past several years have evolved, but continue today ...
In our latest research, Cisco's The App Attention Index 2023: Beware the Application Generation, 62% of consumers report their expectations for digital experiences are far higher than they were two years ago, and 64% state they are less forgiving of poor digital services than they were just 12 months ago ...
In MEAN TIME TO INSIGHT Episode 5, Shamus McGillicuddy, VP of Research, Network Infrastructure and Operations, at EMA discusses the network source of truth ...
A vast majority (89%) of organizations have rapidly expanded their technology in the past few years and three quarters (76%) say it's brought with it increased "chaos" that they have to manage, according to Situation Report 2024: Managing Technology Chaos from Software AG ...
In 2024 the number one challenge facing IT teams is a lack of skilled workers, and many are turning to automation as an answer, according to IT Trends: 2024 Industry Report ...
Organizations are continuing to embrace multicloud environments and cloud-native architectures to enable rapid transformation and deliver secure innovation. However, despite the speed, scale, and agility enabled by these modern cloud ecosystems, organizations are struggling to manage the explosion of data they create, according to The state of observability 2024: Overcoming complexity through AI-driven analytics and automation strategies, a report from Dynatrace ...
Organizations recognize the value of observability, but only 10% of them are actually practicing full observability of their applications and infrastructure. This is among the key findings from the recently completed Logz.io 2024 Observability Pulse Survey and Report ...
Businesses must adopt a comprehensive Internet Performance Monitoring (IPM) strategy, says Enterprise Management Associates (EMA), a leading IT analyst research firm. This strategy is crucial to bridge the significant observability gap within today's complex IT infrastructures. The recommendation is particularly timely, given that 99% of enterprises are expanding their use of the Internet as a primary connectivity conduit while facing challenges due to the inefficiency of multiple, disjointed monitoring tools, according to Modern Enterprises Must Boost Observability with Internet Performance Monitoring, a new report from EMA and Catchpoint ...
