4 citations · 4 across the 7 of their papers we have counts for
6 papers · 1 filter
LLM Evolutionary Search Strategies Should be Compared by Tokens
Tal Oved, Roi Pony, Oshri Naparstek +1
LLM evolutionary search is currently evaluated by counting proposals (candidate programs). We demonstrate this metric hides the true cost of search, and redo the comparison using t…
Spotlights: Discovering Improvement Opportunities in Software Repositories
Udi Barzelay, Ophir Azulai, Idan Friedman +9
Coding agents and evolutionary code-search systems can improve implementations once a target and evaluation criterion have been specified. Applying these methods to an existing sof…
Evolution or Illusion? Rethinking Evaluation in LLM Evolutionary Search
Tal Oved, Roi Pony, Oshri Naparstek +1
LLM-driven evolutionary search finds programs by launching seeds and iterating each one. Papers report a single budget setting, usually one seed run for a fixed number of iteration…
Optimize Cheap, Deploy Strong: Cost-Aware Cross-Tier Transfer for Evolutionary Optimization
Tal Oved, Roi Pony, Oshri Naparstek +1
Evolutionary optimization of LLM prompts and agentic programs (e.g., GEPA) is dominated by fitness evaluation: scoring each candidate runs an answering LLM over a validation set, s…
FLASH-MAXSIM: IO-Aware Fused Kernels for Late-Interaction Retrieval
Roi Pony, Daniel Ezer, Adi Raz Goldfarb +3
Late-interaction retrieval (ColBERT, ColPali) scores a query against a document via the MaxSim operator. The standard PyTorch implementation materialises the full query-token $\tim…
Col-Bandit: Query-Time Top- Estimation for Late-Interaction Retrieval
Roi Pony, Adi Raz Goldfarb, Oshri Naparstek +3
Multi-vector late-interaction retrievers such as ColBERT achieve state-of-the-art quality, but their query-time cost is dominated by exhaustively computing token-level MaxSim inter…