AI Search
Rerankers fine-tuned pointwise on relevant and not-relevant labels have a gap: every negative for a query carries the same label, so the model never learns which negatives sit closer to the positive.
AI Search
Decision models such as Jev generate no text and output a probability for each relevance level directly, and the vendor promises those probabilities are calibrated, yet earlier independent checks never compared them with graded human judgments.
AI Search
GraphRAG in production has two costs that papers rarely count: what it takes to build the graph, and how painful it is to delete a document.
Personalization
On recommendation platforms, large models for different stages and business lines each encode the same long user history, and this repeated cost gets harder to justify as models and sequences multiply.
Personalization
Implicit feedback mixes in accidental clicks and impulse purchases, and existing denoising methods almost all use training loss to find noise, but loss only grows after the model has been misled and is also disturbed by training stage and negative sampling.
Personalization
Airbnb wants to choose cover photos for hosts, and there are two difficulties.
Personalization
When sponsored search ranking uses long-term behavior, the usual approach searches history for actions similar to the candidate product, which can miss history highly relevant to the current query but only loosely tied to the candidate.
Infra
Generative recommendation has the model generate an item's SID token by token like an LLM, so LLM serving stacks look reusable, yet they run poorly because they assume every step has the same shape and attention cost sits in long sequences.
Infra
Agents and RL training issue thousands of GPU jobs that run for seconds, yet starting a Pod takes twenty to thirty seconds even with the image cached, so waiting outweighs the work.
Infra
The same request can produce different output bits just because the other requests in its batch changed, so evaluations cannot be reproduced and an RL trainer recomputing a rollout cannot match it.
Generative
In food delivery search, a merchant that cannot reach the user's current location is not merely ranked low, it must not appear at all.
Generative
Generative recommendation numbers products with SIDs, and the dominant approach is residual quantization: each level has one global code table and looks up whatever the previous level left unexplained, which amounts to assuming that each token carries its own independent meaning.
Generative
On devices such as smart glasses, users often give only a picture and one vague sentence, leaving style, budget and venue requirements unsaid.
Generative
Generative retrieval often has to honor a condition, such as returning only products from one shop, while a trie only guarantees that a generated SID exists and says nothing about shops or countries.
Embodied AI
In two-armed robot demonstrations, which hand moves first is often just the operator's habit, yet imitation learning cannot separate the orderings a task requires from the ones that happened by chance, so when one arm is paused the other waits too.