VQV Signal
Researchers distilled DeepSeek V4 Flash into GPT-OSS-120B for finance tasks, achieving an 83.61% score on FinanceReasoning, outperforming competitors. They found that the distillation process did not transfer the censorship characteristics of the teacher model.
Researchers distilled DeepSeek V4 Flash into GPT-OSS-120B for finance tasks, achieving an 83.61% score on FinanceReasoning, outperforming competitors. They found that the distillation process did not transfer the censorship characteristics of the teacher model.
AI-assisted summary based on the listed source.
We recently used DeepSeek V4 Flash as a teacher for finance tasks with GPT-OSS-120B. Distillation works well on this problem. At a constrained 8k token budget, our self-distilled 120B scores 83.61% on FinanceReasoning, above Kimi K3 (81.93%) and Inkling (65.13%). We released the 20B open weights. With V4 as the...
This shows that high-performance open-source LLMs can be developed via distillation without inheriting restrictive censorship behaviors. It highlights a path to creating more capable and less constrained open models for specialized tasks.
VQV organizes public signals from inspectable sources. It does not independently verify the underlying report.
Signal Strength reflects source quality, relevance, freshness and evidence. Public Interest helps organize discovery; it is not proof of truth.
VQV surfaced this signal because it is recent, relevant to Open Source LLMs, connected to Hacker News Front Page.
No login, cookies, social SDKs, or automatic posting.