Close Menu

    Subscribe to Updates

    Get the latest creative news from FooBar about art, design and business.

    What's Hot

    Teens say they are turning to AI for advice, friendship and ‘to get out of thinking’

    July 23, 2025

    Zelenskyy faces major anti-corruption protests as Ukraine prepares for Russia talks

    July 23, 2025

    Jamia Hamdard’s 150 MBBS, 49 PG seats scrapped for 2025-26; CAG had flagged Rs 813-crore diversion, ETHealthworld

    July 23, 2025
    Facebook X (Twitter) Instagram
    • Demos
    • Buy Now
    Facebook X (Twitter) Instagram YouTube
    14 Trends14 Trends
    Demo
    • Home
    • Features
      • View All On Demos
    • Buy Now
    14 Trends14 Trends
    Home » Unpacking the bias of large language models | MIT News
    Aritifical Intelligence

    Unpacking the bias of large language models | MIT News

    adminBy adminJune 18, 2025No Comments6 Mins Read0 Views
    Facebook Twitter Pinterest LinkedIn Telegram Tumblr Email
    Share
    Facebook Twitter LinkedIn Pinterest Email



    Research has shown that large language models (LLMs) tend to overemphasize information at the beginning and end of a document or conversation, while neglecting the middle.

    This “position bias” means that, if a lawyer is using an LLM-powered virtual assistant to retrieve a certain phrase in a 30-page affidavit, the LLM is more likely to find the right text if it is on the initial or final pages.

    MIT researchers have discovered the mechanism behind this phenomenon.

    They created a theoretical framework to study how information flows through the machine-learning architecture that forms the backbone of LLMs. They found that certain design choices which control how the model processes input data can cause position bias.

    Their experiments revealed that model architectures, particularly those affecting how information is spread across input words within the model, can give rise to or intensify position bias, and that training data also contribute to the problem.

    In addition to pinpointing the origins of position bias, their framework can be used to diagnose and correct it in future model designs.

    This could lead to more reliable chatbots that stay on topic during long conversations, medical AI systems that reason more fairly when handling a trove of patient data, and code assistants that pay closer attention to all parts of a program.

    “These models are black boxes, so as an LLM user, you probably don’t know that position bias can cause your model to be inconsistent. You just feed it your documents in whatever order you want and expect it to work. But by understanding the underlying mechanism of these black-box models better, we can improve them by addressing these limitations,” says Xinyi Wu, a graduate student in the MIT Institute for Data, Systems, and Society (IDSS) and the Laboratory for Information and Decision Systems (LIDS), and first author of a paper on this research.

    Her co-authors include Yifei Wang, an MIT postdoc; and senior authors Stefanie Jegelka, an associate professor of electrical engineering and computer science (EECS) and a member of IDSS and the Computer Science and Artificial Intelligence Laboratory (CSAIL); and Ali Jadbabaie, professor and head of the Department of Civil and Environmental Engineering, a core faculty member of IDSS, and a principal investigator in LIDS. The research will be presented at the International Conference on Machine Learning.

    Analyzing attention

    LLMs like Claude, Llama, and GPT-4 are powered by a type of neural network architecture known as a transformer. Transformers are designed to process sequential data, encoding a sentence into chunks called tokens and then learning the relationships between tokens to predict what words comes next.

    These models have gotten very good at this because of the attention mechanism, which uses interconnected layers of data processing nodes to make sense of context by allowing tokens to selectively focus on, or attend to, related tokens.

    But if every token can attend to every other token in a 30-page document, that quickly becomes computationally intractable. So, when engineers build transformer models, they often employ attention masking techniques which limit the words a token can attend to.

    For instance, a causal mask only allows words to attend to those that came before it.

    Engineers also use positional encodings to help the model understand the location of each word in a sentence, improving performance.

    The MIT researchers built a graph-based theoretical framework to explore how these modeling choices, attention masks and positional encodings, could affect position bias.

    “Everything is coupled and tangled within the attention mechanism, so it is very hard to study. Graphs are a flexible language to describe the dependent relationship among words within the attention mechanism and trace them across multiple layers,” Wu says.

    Their theoretical analysis suggested that causal masking gives the model an inherent bias toward the beginning of an input, even when that bias doesn’t exist in the data.

    If the earlier words are relatively unimportant for a sentence’s meaning, causal masking can cause the transformer to pay more attention to its beginning anyway.

    “While it is often true that earlier words and later words in a sentence are more important, if an LLM is used on a task that is not natural language generation, like ranking or information retrieval, these biases can be extremely harmful,” Wu says.

    As a model grows, with additional layers of attention mechanism, this bias is amplified because earlier parts of the input are used more frequently in the model’s reasoning process.

    They also found that using positional encodings to link words more strongly to nearby words can mitigate position bias. The technique refocuses the model’s attention in the right place, but its effect can be diluted in models with more attention layers.

    And these design choices are only one cause of position bias — some can come from training data the model uses to learn how to prioritize words in a sequence.

    “If you know your data are biased in a certain way, then you should also finetune your model on top of adjusting your modeling choices,” Wu says.

    Lost in the middle

    After they’d established a theoretical framework, the researchers performed experiments in which they systematically varied the position of the correct answer in text sequences for an information retrieval task.

    The experiments showed a “lost-in-the-middle” phenomenon, where retrieval accuracy followed a U-shaped pattern. Models performed best if the right answer was located at the beginning of the sequence. Performance declined the closer it got to the middle before rebounding a bit if the correct answer was near the end.

    Ultimately, their work suggests that using a different masking technique, removing extra layers from the attention mechanism, or strategically employing positional encodings could reduce position bias and improve a model’s accuracy.

    “By doing a combination of theory and experiments, we were able to look at the consequences of model design choices that weren’t clear at the time. If you want to use a model in high-stakes applications, you must know when it will work, when it won’t, and why,” Jadbabaie says.

    In the future, the researchers want to further explore the effects of positional encodings and study how position bias could be strategically exploited in certain applications.

    “These researchers offer a rare theoretical lens into the attention mechanism at the heart of the transformer model. They provide a compelling analysis that clarifies longstanding quirks in transformer behavior, showing that attention mechanisms, especially with causal masks, inherently bias models toward the beginning of sequences. The paper achieves the best of both worlds — mathematical clarity paired with insights that reach into the guts of real-world systems,” says Amin Saberi, professor and director of the Stanford University Center for Computational Market Design, who was not involved with this work.

    This research is supported, in part, by the U.S. Office of Naval Research, the National Science Foundation, and an Alexander von Humboldt Professorship.



    Source link

    Share. Facebook Twitter Pinterest LinkedIn Tumblr Email
    admin
    • Website

    Related Posts

    Beyond accelerators: Lessons from building foundation models on AWS with Japan’s GENIAC program

    July 23, 2025

    How Financial Services Companies Use Agentic AI to Enhance Productivity, Efficiency and Security

    July 22, 2025

    School of Architecture and Planning recognizes faculty with academic promotions in 2025 | MIT News

    July 22, 2025

    Build an AI-powered automated summarization system with Amazon Bedrock and Amazon Transcribe using Terraform

    July 22, 2025

    A new way to edit or generate images | MIT News

    July 21, 2025

    MIT Learn offers “a whole new front door to the Institute” | MIT News

    July 21, 2025
    Leave A Reply Cancel Reply

    Demo
    Top Posts

    ChatGPT’s viral Studio Ghibli-style images highlight AI copyright concerns

    March 28, 20254 Views

    Best Cyber Forensics Software in 2025: Top Tools for Windows Forensics and Beyond

    February 28, 20253 Views

    An ex-politician faces at least 20 years in prison in killing of Las Vegas reporter

    October 16, 20243 Views

    Laws, norms, and ethics for AI in health

    May 1, 20252 Views
    Don't Miss

    Teens say they are turning to AI for advice, friendship and ‘to get out of thinking’

    July 23, 2025

    No question is too small when Kayla Chege, a high school student in Kansas, is…

    Zelenskyy faces major anti-corruption protests as Ukraine prepares for Russia talks

    July 23, 2025

    Jamia Hamdard’s 150 MBBS, 49 PG seats scrapped for 2025-26; CAG had flagged Rs 813-crore diversion, ETHealthworld

    July 23, 2025

    Trump rehashes years-old grievances on Russia investigation after new intelligence report

    July 23, 2025
    Stay In Touch
    • Facebook
    • Twitter
    • Pinterest
    • Instagram
    • YouTube
    • Vimeo

    Subscribe to Updates

    Get the latest creative news from SmartMag about art & design.

    Demo
    Top Posts

    ChatGPT’s viral Studio Ghibli-style images highlight AI copyright concerns

    March 28, 20254 Views

    Best Cyber Forensics Software in 2025: Top Tools for Windows Forensics and Beyond

    February 28, 20253 Views

    An ex-politician faces at least 20 years in prison in killing of Las Vegas reporter

    October 16, 20243 Views
    Stay In Touch
    • Facebook
    • YouTube
    • TikTok
    • WhatsApp
    • Twitter
    • Instagram
    Latest Reviews
    Demo
    About Us
    About Us

    Your source for the lifestyle news. This demo is crafted specifically to exhibit the use of the theme as a lifestyle site. Visit our main page for more demos.

    We're accepting new partnerships right now.

    Email Us: info@example.com
    Contact: +1-320-0123-451

    Facebook X (Twitter) Pinterest YouTube WhatsApp
    Our Picks

    Teens say they are turning to AI for advice, friendship and ‘to get out of thinking’

    July 23, 2025

    Zelenskyy faces major anti-corruption protests as Ukraine prepares for Russia talks

    July 23, 2025

    Jamia Hamdard’s 150 MBBS, 49 PG seats scrapped for 2025-26; CAG had flagged Rs 813-crore diversion, ETHealthworld

    July 23, 2025
    Most Popular

    ChatGPT’s viral Studio Ghibli-style images highlight AI copyright concerns

    March 28, 20254 Views

    Best Cyber Forensics Software in 2025: Top Tools for Windows Forensics and Beyond

    February 28, 20253 Views

    An ex-politician faces at least 20 years in prison in killing of Las Vegas reporter

    October 16, 20243 Views

    Subscribe to Updates

    Get the latest creative news from FooBar about art, design and business.

    14 Trends
    Facebook X (Twitter) Instagram Pinterest YouTube Dribbble
    • Home
    • Buy Now
    © 2025 ThemeSphere. Designed by ThemeSphere.

    Type above and press Enter to search. Press Esc to cancel.