Tech Research (10)

Part of the Technology feed · AI-researched.

Latest

HypoKG: Evidence-Grounded Biomedical Hypothesis Generation

System that generates biomedical hypotheses grounded in scientific evidence from major biological databases rather than producing unsupported ideas.

SynthSentry: Detecting Synthetic Data Contamination in LLM Training

Method to identify synthetic data contamination in language model training data before training occurs, addressing model collapse prevention.

Agent as Policy: General-Purpose Agents Driving Robotic Manipulation

Demonstrates that a general-purpose agent can directly control a physical robot for task execution without task-specific or environment-specific training.

Research: LLMs can infer sensitive attributes from aggregated social media posts

A study shows language models can accurately infer age, income, and occupation from aggregated user-generated content via personal knowledge graphs.

R2VC: modular fact-checking framework combining retrieval, verification and calibration

Researchers introduce R2VC, a modular fact-checking system that separates evidence retrieval, reasoning, and confidence estimation to improve LLM reliability.

Rate-distortion analysis of factual hallucination in closed-book QA

Research demonstrates that factual hallucination in LLMs stems partly from compression trade-offs, even when facts are present in internal representations.

Automated Detection of Social Tipping Points in Climate Literature

AI framework automatically identifies and structures evidence of climate-related social tipping points across the climate literature.

Adaptive Self-Correcting Inference for Post-ASR False Wake-Up Prevention

Method to reduce false wake-up activations in voice assistants when phonetically similar speech is misrecognized as the wake word.

AMDKernelVault: Open Datasets and Training for AMD GPU Kernel Optimization

Open HIP and Triton kernel corpus and training framework for optimizing code on recent AMD CDNA GPUs, addressing NVIDIA-centric limitations.

Meddies-PII: Multilingual PII Extraction Framework for Clinical De-identification

Framework for extracting personally identifiable information in clinical documents across multiple languages, addressing dataset scarcity challenges.

Earlier

2026-09-12 (10 events) · 2026-09-11 (10 events) · 2026-09-09 (10 events) · 2026-09-08 (1 event) · 2026-09-07 (10 events) · 2026-09-05 (10 events) · 2026-09-03 (10 events) · 2026-09-02 (10 events) · 2026-09-01 (10 events) · 2026-08-31 (10 events) · 2026-08-29 (10 events) · 2026-08-28 (10 events) · 2026-08-27 (10 events) · 2026-08-26 (10 events) · 2026-08-25 (10 events) · 2026-08-24 (10 events) · 2026-08-23 (10 events) · 2026-08-22 (4 events) · 2026-08-21 (10 events) · 2026-08-20 (10 events) · 2026-08-19 (10 events) · 2026-08-18 (10 events) · 2026-08-17 (10 events) · 2026-08-16 (1 event) · 2026-08-15 (10 events) · 2026-08-14 (10 events) · 2026-08-13 (10 events) · 2026-08-12 (10 events) · 2026-08-11 (3 events) · 2026-08-10 (3 events) · 2026-08-09 (3 events) · 2026-08-08 (2 events) · 2026-08-07 (2 events) · 2026-08-06 (2 events) · 2026-08-05 (3 events) · 2026-08-04 (2 events) · 2026-08-03 (3 events) · 2026-08-02 (3 events) · 2026-08-01 (1 event) · 2026-07-31 (1 event)