SponsoredChatgpt/Claude/Grok 会员24h在线自助充值 安全快速 稳定保障
奥创(Ultra)商城会员充值中心
SponsoredChatgpt/Claude/Grok 会员24h在线自助充值 安全快速 稳定保障
奥创(Ultra)商城会员充值中心
Sponsored本页公开展示实时样本、模型覆盖与响应时间。先用 5 元按量跑自己的 Claude Code/Codex 任务,成功后再决定是否长期使用。
ACS Gateway
Sponsored【腾讯云】WorkBuddy 全场景 AI 办公工作台,团队旗舰版低至198元/月
腾讯云
Sponsored人人可用的AI应用创作平台 自然语言交互,全栈应用搭建,伟大创意快速实现;个人版新用户注册送5000积分
阿里云OpenAI says Apollo Research, METR, Transluce, and Goodfire will independently examine aspects of its models, evaluations, and safeguards.
The first phase of OpenAI’s embedded assessments gives outside experts deep access to the company’s internal models and systems. OpenAI says the work builds on its existing alignment and safety research and is intended to test how risks are identified and managed.
The groups will investigate different areas: Apollo Research will examine whether models might act covertly against developers’ intentions; METR will study potential misaligned behavior in a major reinforcement-learning run; Transluce will review post-training alignment evaluations; and Goodfire will analyze model internals and related monitoring and debugging work.
Source review does not guarantee permanent accuracy. Consider the original source, observation time, and later corrections.
Key points6 · 5 min full read 1. Four small figures at an open door. On November 16, 2026, OpenAI announced embedded assessments: Apollo Research, METR, Transluce and Goodfire get deep access to its internal models and systems to test alignment and safeguards. 2. Two stacked reports with a magnifying glass. Apollo Research will publish two reports. For its training assessment OpenAI committed code, fine-tuning data, RL rollouts with reasoning traces and rewards, and checkpoints it can query. 3. A single loop arrow inside a fenced box. METR’s pilot covers OpenAI’s largest RL run only. It does not cover the misalignment incidents OpenAI has disclosed so far, which METR calls narrower than it proposed. 4. A calendar with six marked squares. [...] Skip to content CellCog Log inSign upHire an AI employee AI EmployeeSuper-AgentsAgent-to-AgentBenchmarksPricingBlog Log inSign up Guides 07 Trust, permissions & security # OpenAI Lets Four Outside Groups Test Its Models Nitish
OpenAI named four outside groups: Apollo Research, METR, Transluce, and Goodfire. These groups get deep access to OpenAI's internal models and systems to investigate alignment and test the company's safeguards, as part of a program OpenAI calls embedded assessments, building on its existing alignment and safety research. Developers tracking AI safety practices follow updates like this one on daily.dev. Share this post
Quartz Business News # OpenAI is bringing four independent safety evaluators inside its systems The pilot assessments cover model alignment, safeguards and monitoring, and give outside experts access to OpenAI's code and training data By Ella·2 min read·Updated November 16, 2026 Add QZ to Google SOPA Images / Getty Images OpenAI signed agreements to bring in four outside assessors — Apollo Research, METR, Transluce and Goodfire — for the first phase of its embedded safety assessments, the company said on Monday. [...] Apollo Research will assess whether models could secretly act in opposition to what their developers intend, and whether OpenAI's safeguards would catch such conduct. Goodfire will embed interpretability researchers to analyze model internals, concentrating on how training runs are monitored, along with debugging and alignment work. Goodfire will publish findings as the collaboration progresses.
@openai: @TransluceAI will assess our post-training alignment evaluations, examining their quality and coverage. They will develop additional tests and recommend improvements to address gaps in what we test for, and intend to expand our pilot in 2027. @openai: @ApolloResearch will assess whether our models could covertly work against their developers' intentions—and whether our safeguards would detect it. It will examine evidence from training, safety testing, and internal use to test for this behavior and our ability to catch it. @openai: @GoodfireAI will embed interpretability researchers to use their platform to analyze our models' internals. Their work focuses on training run monitoring, debugging, and alignment, helping our researchers identify and make safety improvements. [...] @goodfireai: Announcing our collaboration with OpenAI Marius Hobbhahn / @mariushobbhahn: Excited about this pilot towards embedded evals with OpenAI! We're specifically focused on understandi
Views 601 @OpenAI OpenAI@OpenAI 11h @ApolloResearch will assess whether our models could covertly work against their developers’ intentions—and whether our safeguards would detect it. It will examine evidence from training, safety testing, and internal use to test for this behavior and our ability to catch it. 20 @OpenAI OpenAI@OpenAI 11h @METR\_Evals will investigate misaligned behavior in our largest reinforcement learning run to date, including failures our existing tests and monitors may have missed. We’ve also preliminarily agreed to conduct a broader monitor assessment, pending mutual agreement on scope. 6 @OpenAI OpenAI@OpenAI 11h [...] 6 @OpenAI OpenAI@OpenAI 11h @TransluceAI will assess our post-training alignment evaluations, examining their quality and coverage. They will develop additional tests and recommend improvements to address gaps in what we test for, and intend to expand our pilot in 2027. 5 @OpenAI OpenAI
Views 5 @OpenAI OpenAI@OpenAI 2h We’re working with @ApolloResearch, @METR\_Evals, @TransluceAI, and @GoodfireAI on embedded assessments: giving outside experts deep access to our internal models and systems to investigate alignment and test our safeguards. These assessments build on our ongoing alignment and safety research, bringing independent scrutiny to how we identify and manage risks in our systems. Our first phase of embedded assessments, in white on green and mint Neutral library artwork. Our first phase of embedded assessmentsFrom openai.com 466 @OpenAI OpenAI@OpenAI 2h [...] 466 @OpenAI OpenAI@OpenAI 2h @ApolloResearch will assess whether our models could covertly work against their developers’ intentions—and whether our safeguards would detect it. It will examine evidence from training, safety testing, and internal use to test for this behavior and our ability to catch it. 11 @OpenAI OpenAI@OpenAI 2h @METR\_Evals will inves