Amazon Science

Reinforcement learning (RL) is one of the techniques we use to make language models better at sustained, multistep tasks like writing code, navigating a website, or carrying out a research workflow. The model doesn't act alone in those settings; it's wrapped in a piece of software we call a harness, which lets it call tools, observe the results of using them, and decide what to do next.

Tabular foundation models like TabPFN and TabICL achieve state-of-the-art performance through in-context learning, yet their architectures remain fundamentally opaque. We introduce KernelICL, a framework to enhance tabular foundation models with quantifiable sample-based inspectability.

Expo Talk: Strands robots: Unifying robot control, simulation, and training behind natural language July 7, 12:00 AM - 1:00 AM KST Room: Hall B2 Speakers: Cagatay Cali, Yin Song Abstract: Despite rapid advances in vision-language-action (VLA) models, deploying robot intelligence remains fragmented: different SDKs for different robots, different policy frameworks with incompatible interfaces, an unbridged simulation-to-reality gap, and training pipelines that demand specialist expertise .

At Amazon, we believe that as our strategy to reach net-zero by 2040 evolves, we need to continue to raise the bar on what and how we measure. Measuring carbon emissions across our entire business is complex, and the tools and methodologies available are continuously improving.

Multi-Agent Debate (MAD) frameworks improve factual reliability in large language models (LLMs) by allowing agents to critique and refine one another's reasoning. Yet, existing MAD systems are computationally expensive and prone to degradation under prolonged debates due to redundant exchanges and unstable judging.

Goal-oriented dialog systems enable users to complete specific goals like requesting information about a movie or booking a ticket. Typically the dialog system pipeline contains multiple ML models, including natural language understanding, state tracking and action prediction (policy learning). These models are trained through a combination of supervised or reinforcement learning methods and therefore require collection of labeled domain specific datasets.

Deploying LLM-based analytics agents in enterprise settings requires evaluation frameworks that can reliably detect failures across complex, multi-tool workflows. We present a three-phase comparative study of three evaluation frameworks (Strands Evals, PromptFoo, and Agenta) applied to two analytics agents in a controlled research setting using frozen execution traces.

Mitigating the retention of sensitive or private information in large language models is essential for enhancing privacy and safety. Existing unlearning methods, like Gradient Ascent and Negative Preference Optimization, directly tune models to remove unwanted information. However, these methods often become unstable because they fine-tune by maximizing cross-entropy loss, which is the opposite of traditional loss minimization in learning.

E-commerce companies like Amazon, Alibaba and Flipkart have an extensive catalogue comprising of billions of products. Matching customer search queries to plausible products is challenging due to the size and diversity of the catalogue. These challenges are compounded in apparel due to the semantic complexity and a large variation of fashion styles, product attributes and colours.

In this paper we introduce SemiGPC, a distribution-aware label refinement strategy based on Gaussian Processes where the predictions of the model are derived from the labels posterior distribution. Differently from other buffer-based semi-supervised methods such as Co-Match [17] and SimMatch [34], our SemiGPC includes a normalization term that addresses imbalances in the global data distribution while maintaining local sensitivity.