PulseAugur
EN
LIVE 09:11:48

HKUST's HarnessSQL boosts 8B model to 45.2% on Spider benchmark

By PulseAugur Editorial · [1 sources] ·

Researchers at HKUST have developed HarnessSQL, a system that trains smaller language models within the DeepSeek Harness framework. This approach successfully boosted an 8 billion parameter model's performance on the Spider 2.0-SQLite benchmark from 15.5% to 45.2%, surpassing larger models that scored around 16%. The technique focuses on efficiently enhancing the capabilities of smaller models for text-to-SQL tasks. AI

IMPACT Demonstrates an effective method for improving the performance of smaller language models on complex tasks like text-to-SQL.

RANK_REASON Research milestone achieved by a university lab on a specific benchmark. [lever_c_demoted from research: ic=1 ai=1.0]

Read on Pandaily →

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →

HKUST's HarnessSQL boosts 8B model to 45.2% on Spider benchmark

How we ranked this

Signal score
8 / 100
Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket
Tool
Research milestone achieved by a university lab on a specific benchmark. [lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration
Single-source cluster
Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics
model release, product
Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance
High
Clearly on-topic for AI-industry coverage.
Story freshness
Same-day
Cluster formed today. Ranking reflects the current source set at time of score.

Full methodology in our editorial standards.

COVERAGE [1]

  1. Pandaily TIER_1 English(EN) · [email protected] (Pandaily) ·

    HarnessSQL Uses DeepSeek Harness to Lift an 8B Model to 45.2% on Spider 2.0-SQLite

    HKUST's HarnessSQL trains small models inside DeepSeek Harness, lifting an 8B model from 15.5% to 45.2% on Spider 2.0-SQLite and past 32B baselines that score about 16%.

RELATED ENTITIES

RELATED TOPICS