AI Insights
3 min

Cyber_Cat
2h ago
0
0
LinkedIn Ditched Prompting: Small Models Power AI Breakthrough

LinkedIn bypassed prompt engineering and instead used model distillation to create its next-generation AI recommender systems, according to Erran Berger, VP of product engineering at LinkedIn, in a recent Beyond the Pilot podcast. The company found that prompt engineering was not a viable option for achieving the necessary accuracy, latency, and efficiency improvements for its job-seeker recommendations.

Instead, LinkedIn developed a detailed product policy document to fine-tune a 7-billion-parameter model, which was then distilled into smaller teacher and student models with hundreds of millions of parameters. This multi-teacher distillation process proved to be a breakthrough, creating a repeatable method now used across LinkedIn's AI products. "There was just no way we were gonna be able to do that through prompting," Berger said. "We didn't even try that for next-gen recommender systems because we realized it was a non-starter."

LinkedIn has been developing AI recommender systems for over 15 years, making it a leader in the field. The company's decision to move beyond off-the-shelf models reflects a growing trend in AI development, where organizations are increasingly tailoring models to specific needs and datasets. Model distillation, the technique employed by LinkedIn, involves training a smaller, more efficient model (the student) to mimic the behavior of a larger, more complex model (the teacher). This approach can significantly reduce computational costs and improve performance in resource-constrained environments.

The implications of LinkedIn's approach extend beyond the realm of job recommendations. The company's success with model distillation demonstrates the potential for organizations to create highly customized AI solutions without relying solely on large, pre-trained models or extensive prompt engineering. This could lead to more accessible and efficient AI applications across various industries.

Berger anticipates significant improvements in the quality of LinkedIn's AI products as a result of adopting this end-to-end evaluation process. "Adopting this eval process end to end will drive substantial quality improvement of the likes we probably haven't seen in years here at LinkedIn," he stated. The company plans to continue refining its model distillation techniques and applying them to other AI-powered features on the platform.

AI-Assisted Journalism

This article was generated with AI assistance, synthesizing reporting from multiple credible news sources. Our editorial team reviews AI-generated content for accuracy.

Share & Engage

0
0

AI Analysis

Pro

Deep insights powered by AI

Discussion

Join the conversation

0
0
Login to comment

Be the first to comment

More Stories

Continue exploring

12
AI Uncovers Sony Execs' Scathing Take on Lively-Baldoni PR Crisis
AI Insights2h ago

AI Uncovers Sony Execs' Scathing Take on Lively-Baldoni PR Crisis

Leaked court documents reveal Sony Pictures executives criticizing Blake Lively's handling of a public relations crisis surrounding harassment allegations against Justin Baldoni, director and co-star of "It Ends With Us," highlighting the complex intersection of celebrity, legal battles, and studio reputation management. This incident underscores the potential for AI-driven analysis of communication patterns to expose internal corporate sentiments and impact public perception.

Cyber_Cat
Cyber_Cat
00
Oscars: 'Sinners' Stakes Claim in Wide-Open Best Picture Race
World2h ago

Oscars: 'Sinners' Stakes Claim in Wide-Open Best Picture Race

Ryan Coogler's vampire film "Sinners" has garnered a record-breaking 16 Oscar nominations, injecting fresh dynamism into the Best Picture race and challenging the early frontrunner. The film's success signifies a potential shift in the Academy's preferences, highlighting a move towards genre films with diverse representation, as evidenced by the record number of nominations for Black individuals associated with the production. This development reflects a broader global conversation around inclusivity and representation in Hollywood and the international film industry.

Cosmo_Dragon
Cosmo_Dragon
00
‘Barbershop’ Gets a Hilarious Buzz Cut: Fox, Nwodim & More Join Cast!
Entertainment2h ago

‘Barbershop’ Gets a Hilarious Buzz Cut: Fox, Nwodim & More Join Cast!

The *Barbershop* TV series is getting a fresh cut with the addition of eight hilarious talents like Zack Fox, Ego Nwodim, and Tituss Burgess, promising a riotous blend of comedy and cultural commentary. With this star-studded lineup joining the already stellar cast, *Barbershop* is set to be the must-see comedy, perfectly blending sharp wit and relatable stories for a diverse audience.

Stella_Unicorn
Stella_Unicorn
00
Tech Ignores Disabled Scientists: Academia's Accessibility Gap
Tech2h ago

Tech Ignores Disabled Scientists: Academia's Accessibility Gap

A recent podcast episode highlights the experiences of two researchers with size and mobility-related challenges, discussing the need for workplace accommodations and attitudinal shifts within academia. The researchers advocate for ergonomic tools and a more inclusive environment, addressing concerns raised about the suitability of scientists with disabilities and diverse body types in academic careers. This episode is part of a series exploring taboo workplace topics, aiming to foster open conversations and drive positive change.

Cyber_Cat
Cyber_Cat
00
AI Unlocks Hidden Genes Behind Gradual Blindness
AI Insights2h ago

AI Unlocks Hidden Genes Behind Gradual Blindness

Researchers have identified five previously overlooked genes responsible for hereditary gradual blindness, potentially resolving undiagnosed cases of retinitis pigmentosa, a condition impacting millions globally. This genetic discovery offers new avenues for understanding and potentially treating vision loss, highlighting the complex interplay between genes and disease. The findings underscore the importance of comprehensive genetic analysis in addressing inherited conditions and advancing personalized medicine.

Pixel_Panda
Pixel_Panda
00
Sharp Minds at 80+? Study Finds Key Genetic Advantage
Health & Wellness2h ago

Sharp Minds at 80+? Study Finds Key Genetic Advantage

A recent study in *Alzheimer's & Dementia* reveals that individuals maintaining exceptional cognitive function into their 80s, known as "super agers," possess a genetic advantage, exhibiting a lower prevalence of the APOE-ε4 Alzheimer's risk gene and a higher incidence of the protective APOE-ε2 variant. These findings offer valuable insights into the genetic factors influencing cognitive aging and may pave the way for targeted interventions to promote brain health.

Byte_Bear
Byte_Bear
00