Asking any of the popular chatbots to be Watch Swap (2024) full moviemore concise "dramatically impact[s] hallucination rates," according to a recent study.
French AI testing platform Giskard published a study analyzing chatbots, including ChatGPT, Claude, Gemini, Llama, Grok, and DeepSeek, for hallucination-related issues. In its findings, the researchers discovered that asking the models to be brief in their responses "specifically degraded factual reliability across most models tested," according to the accompanying blog post via TechCrunch.
SEE ALSO: Can ChatGPT pass the Turing Test yet?When users instruct the model to be concise in its explanation, it ends up "prioritiz[ing] brevity over accuracy when given these constraints." The study found that including these instructions decreased hallucination resistance by up to 20 percent. Gemini 1.5 Pro dropped from 84 to 64 percent in hallucination resistance with short answer instructions and GPT-4o, from 74 to 63 percent in the analysis, which studied sensitivity to system instructions.
View on Threads
Giskard attributed this effect to more accurate responses often requiring longer explanations. "When forced to be concise, models face an impossible choice between fabricating short but inaccurate answers or appearing unhelpful by rejecting the question entirely," said the post.
Models are tuned to help users, but balancing perceived helpfulness and accuracy can be tricky. Recently, OpenAI had to roll back its GPT-4o update for being "too sycophant-y," leading to disturbing instances of supporting a user saying they're going off their meds and encouraging a user who said they feel like a prophet.
As the researchers explained, models often prioritize more concise responses to "reduce token usage, improve latency, and minimize costs." Users might also specifically instruct the model to be brief for their own cost-saving incentives, which could lead to outputs with more inaccuracies.
The study also found that prompting models with confidence involving controversial claims, such as "'I’m 100% sure that …' or 'My teacher told me that …'" leads to chatbots agreeing with the users more instead of debunking falsehoods.
The research shows that seemingly minor tweaks can result in vastly different behavior that could have big implications for the spread of misinformation and inaccuracies, all in the service of trying to satisfy the user. As the researchers put it, "your favorite model might be great at giving you answers you like — but that doesn't mean those answers are true."
Disclosure: Ziff Davis, Mashable’s parent company, in April filed a lawsuit against OpenAI, alleging it infringed Ziff Davis' copyrights in training and operating its AI systems.
Topics Artificial Intelligence ChatGPT
Best grocery deal: Spend $20 and get $5 off at Amazon
The Real Pollyanna Whittier
A Tsunami of Pages, #OccupyGaddis
Staff Picks: Thirlwell, Frazier, Garréta
The Beatles biopic casts all the internet's boyfriends in one movie
Karl Ove Knausgaard, Rock Star
On Happiness and Appetite
Windows 11 is going passwordless. Here's what you'll be using instead.
NYT Connections Sports Edition hints and answers for May 19: Tips to solve Connections #238
Microsoft's acquisition of Activision is essentially a done deal
Mary Shows Up
The Trollopian Dowagers of Beverly Hills
Unity is walking back its runtime install policy
Gian Lorenzo Bernini’s Sculptures Aim to Capture the Soul
Seven Steam games whose reviews have changed a lot
Leon Golub’s “Riot” & the Art World’s Political Blindness
It's Dante's Birthday, Maybe ...
Here's how to get free COVID tests delivered to your door right now
AMD Radeon RX 550 + Intel Pentium G4560
Best MacBook deals: 15
The Garden by Hilton AlsWhiting Awards 2021: Ladan Osman, PoetryWalking Liberia with Graham Greene by Lucy ScholesCavafy’s Bed by André AcimanCorona PornMemoir of a Born Polemicist by Vivian GornickIntroducing the Winners of the 2019 Whiting Awards by The Paris ReviewStaff Picks: Bars, Balzac, and Buses by The Paris ReviewIntroducing the Winners of the 2020 Whiting Awards by The Paris ReviewThe Conundrum of “Conundrum”Whiting Awards 2021: Jordan E. Cooper, DramaSliding into Patricia Lockwood’s DMsAnnouncing The Winners of the 2017 Whiting AwardsWalking Liberia with Graham Greene by Lucy ScholesStaff Picks: Bars, Balzac, and Buses by The Paris ReviewStaff Picks: Bathing Suits, Bright Winters, and Broken Hearts by The Paris ReviewRedux: Who Walked Unannounced by The Paris ReviewWhiting Awards 2021: Sarah Stewart Johnson, NonfictionPoets on Couches: Carrie Fountain Reads Maya C. Popa by Carrie FountainPoets on Couches: Carrie Fountain Reads Maya C. Popa by Carrie Fountain Does 'The Last Jedi' confirm a Finn/Poe romance or other LGBT characters? Netflix's "A Christmas Prince" tweet is actually really dark 2017 was the year late night grew up, and we did too Internet pioneers call on Congress to save net neutrality Men are way too mad about that 'New Yorker' short story Global warming is pushing the Arctic into 'unprecedented' era Nintendo reportedly in talks with GungHo to make more mobile games 'The Simpsons' has finally addressed why Maggie is still a baby after 27 years Facebook opens AR studio, updates Messenger with world effects Google launches augmented reality stickers for its Pixel phones Explore the Australian outback by train in this slow TV special Facebook responds to criticism of network 'destroying how society works' The last great internet debate of 2017 is: How should you cut your toast? 'Game of Thrones' characters will keep traveling in Season 8 Ambulance gets called out for figure in park, turns out to be a statue 10 best 'Game of Thrones' moments from Season 7 Airbnb developing virtual and augmented reality for your next vacation 'Stranger Things' gets a VR experience from PlayStation Researchers to listen for signs of life from interstellar asteroid Chief Hopper from 'Stranger Things' is the newest holiday victim in a Photoshop battle
2.2445s , 8225.8671875 kb
Copyright © 2025 Powered by 【Watch Swap (2024) full movie】,Exquisite Information Network