Tagged articles
ablation study
10 articles · Page 1 of 1
Researchers from Northwestern Polytechnical University and HKUST propose Dynamic Gated Pathways (DGP), a method that lets each LLM layer dynamically attend to all ViT layers, improving multimodal benchmarks by learning per-layer, per-sample routing weights instead of using a fixed visual projection.
DGPDynamic Gated PathwaysLLM
0 likes · 18 min read

A study from XJTU, SJTU, and ZJU shows RAW-RAG — a simple dense retrieval + reranking pipeline on raw dialogue history — outperforms 12 agentic memory systems across four benchmarks, achieving higher accuracy with 83× faster ingestion because raw evidence preserves reasoning paths that compression loses.
LLMLong-term MemoryRAG
0 likes · 7 min read

Alibaba's QwenPaw-Data splits enterprise data analytics into three layers—DataBridge for facts, Skill-Hub for methodology, and Host for execution—showing that semantic layer alone is insufficient; ablation experiments reveal methodology constraints dramatically improve analysis breadth, depth, and artifact completeness.
AlibabaData AgentQwenPaw-Data
0 likes · 18 min read

EvoQuant introduces a self‑evolving, validator‑guided framework that lets large language models diagnose, propose, and verify improvements to existing quantitative trading strategies, achieving Sharpe ratio gains of up to 199% across A‑share and cryptocurrency benchmarks while demonstrating robustness through ablation and walk‑forward tests.
EvoQuantSharpe ratioablation study
0 likes · 19 min read

In a YC interview, Boris Cherny, the creator of Claude Code, urges AI product teams to repeatedly delete and test system prompts, adopt ablation studies, and embrace concepts like Product Overhang and Unhobbling, offering ten concrete recommendations for building and evaluating generative AI systems.
AI adviceAI productClaude
0 likes · 10 min read

The article critically examines the assumption that longer chain‑of‑thought reasoning always improves large language model performance, presents the E‑GRM framework that dynamically decides when to invoke full CoT based on model‑internal uncertainty, and validates its efficiency and accuracy gains through extensive experiments and ablations.
EfficiencyGenerative Reward Modelablation study
0 likes · 16 min read

The paper introduces Agentic Harness Engineering (AHE), showing that a 10‑round evolution improves Coding Agent pass@1 from 69.7% to 77.0% on Terminal‑Bench 2—outperforming Codex‑CLI—and that the evolved harness transfers zero‑shot to SWE‑bench and multiple model families, thanks to three observability pillars.
Agentic AICoding AgentHarness Engineering
0 likes · 11 min read

This article reviews the KANMixer model, which places Kolmogorov‑Arnold Networks at the core of a lightweight architecture for long‑term time series forecasting, detailing its design, extensive benchmark experiments on seven real‑world datasets, ablation analyses, and its computational trade‑offs versus MLP and Transformer baselines.
KANLong-term Time Series ForecastingMLP
0 likes · 8 min read

This article reviews the MLLM4TS framework, which fuses visual representations of multivariate time series with large language models to address complex temporal dependencies, cross‑channel interactions, and task generalization, and demonstrates superior performance on classification, anomaly detection, forecasting, and few‑shot scenarios across multiple benchmarks.
Few-Shot LearningVision-Language Fusionablation study
0 likes · 11 min read

GPTuner, an LLM‑driven system for PostgreSQL knob tuning developed by researchers at Sichuan University, demonstrates that knowledge processing, parameter selection, search‑range optimization, and a two‑stage Bayesian framework each significantly improve performance, while costing roughly 880 000 GPT‑4 tokens (≈ $30) with reusable knowledge.
GPTunerLLMPostgreSQL
0 likes · 9 min read
