Three Arrows Co-Founder Su Zhu Faces Questioning in Singapore Court in Hunt for Assets: Bloomberg

CoinDeskPolicyPubblicato 2023-12-12Pubblicato ultima volta 2023-12-13

Introduzione

Zhu is expected to be released from jail this month for good behavior, Bloomberg reported.

Su Zhu, a co-founder of failed crypto hedge fund Three Arrows Capital (3AC), faced questions in a Singapore court for the first time as liquidators sought information to help retrieve assets, Bloomberg reported on Wednesday.

The two-day court hearing required Zhu to respond to lawyers for liquidator Teneo with details on how the fund failed and the whereabouts of the assets, people familiar with the matter told Bloomberg. The people wished to remain anonymous because the proceedings were private, the report said.

10

The questioning was approved after Zhu was arrested in September after failing to help wind up 3AC. Zhu, jailed for four months, is expected to be released this month for good behavior, Bloomberg reported.

Advertisement
Advertisement

Three Arrows Capital filed for Chapter 15 bankruptcy in July last year after suffering losses following the collapse of stablecoin issuer Terra. In June, it was reported that liquidators had been seeking $1.3 billion from the bankrupt fund's co-founders.

Teneo did not respond to CoinDesk's request for comment. Su Zhu declined to comment via Telegram.

Ian Allison contributed to reporting.

Edited by Sheldon Reback.

Letture associate

Just now, DeepSeek V4 updates with DSpark, improving inference speed by 80%

DeepSeek has updated its DeepSeek V4 model with the DSpark speculative decoding framework, achieving a significant 60-85% speedup in generation for Flash models and 57-78% for Pro models while maintaining the same overall throughput. This engineering-focused update, rather than a core architectural change, introduces DSpark to address latency and throughput bottlenecks in high-concurrency production environments. DSpark combines high-throughput parallel generation with adaptive load-aware verification. Its key innovations include a semi-autoregressive generation architecture to model dependencies within token blocks and a hardware-aware confidence-scheduled verification system. This system uses a confidence head to predict token acceptance probabilities, allowing it to dynamically optimize verification length per request and allocate compute only to tokens with the highest expected payoff. The asynchronous scheduler is designed for real-world deployment, ensuring zero-overhead scheduling and continuous CUDA graph replay while preserving the target model's output distribution. In tests across mathematical reasoning, code generation, and daily dialogue, DSpark outperformed state-of-the-art models like Eagle3 and DFlash, increasing average acceptance length by 26.7%-30.9% and 16.3%-18.4% respectively on Qwen3 target models. DeepSeek also open-sourced DeepSpec, a full-stack codebase for training and evaluating speculative decoding draft models, providing a standardized toolkit that includes data preparation tools, model implementations, training code, and evaluation scripts.

marsbit8 h fa

Just now, DeepSeek V4 updates with DSpark, improving inference speed by 80%

marsbit8 h fa

Trading

Spot
活动图片