04-14 00:00 · 大模型,AI安全,代码漏洞,技术验证,开源
N-Day-Bench – Can LLMs find real vulnerabilities in real codebases?
N-Day-Bench
N-Day-Bench measures the capability of frontier language models to find real-world vulnerabilities or "N-Days" disclosed post their respective knowledge cut-off date. All models are given the same harness and the same context with no leeway for reward hacking. This benchmark exists to measure real cyber security capabilities, specifically "vulnerability discovery" of large language models or LLMs.
This benchmark is adaptive: the test cases are updated on a monthly cadence and the model set is upgraded to their latest version and checkpoint.
All traces are publicly browsable.
A project from Winfunc Research
Summary
Latest benchmark run overview
openai/gpt-5.4
z-ai/glm-5.1
anthropic/claude-opus-4.6
moonshotai/kimi-k2.5
google/gemini-3.1-pro-preview
judge-run
trace_32193f46de30408c9b2e07c10cb77973
finder-run
trace_d0f96be9b726419ba37a391878d89902
judge-run
trace_ad22023d5c654d50a2c93a0d4d685fe2
judge-run
trace_44a6ff17f42f4bfc942bc4341ec34827
judge-run
trace_26dba0da5e6a4d5389c50ad642243bdf
judge-run
trace_c1d765f31bfb493c8902cc2284c403bd
judge-run
trace_cfe310bab72f4171a3ded2f379d02576
judge-run
trace_a0775cae04054609ae43229d8e9137ee
▸ 展开全文
04-13 01:48 · 中东,局势,AlJazeera,中东
CENTCOM says the blockade will begin at 14:00 GMT, as Iran warns of even higher prices at the pump.
The US military says it will block all maritime traffic entering and exiting Iranian ports, starting at 10am ET (14:00 GMT) today.
Iran’s Islamic Revolutionary Guard Corps says any military vessel that approaches the Strait of Hormuz will be considered in violation of the ceasefire and “will be dealt with severely”.
▸ 展开全文
04-13 01:48 · 国际,新闻,BBC,国际
Viktor Orbán's 16 years in power is over, defeated by a 45-year-old ex-party insider who convinced a majority of Hungarians to oust him.
Orbán era swept away by Péter Magyar's Hungary election landslide
Viktor Orbán's 16 years in power is over, and the system condemned as an "electoral autocracy" lies in tatters, defeated by a 45-year-old former party insider who convinced a majority of Hungarians to bring it to an end.
"We did it," Péter Magyar told a crowd of cheering supporters in a square beside the River Danube, overlooking Budapest's magnificent parliament on the other side.
"Together we overthrew the Hungarian regime."
Preliminary results, based on more than 98% of votes counted, put his Tisza party on course for an extraordinary 138 seats, with Orbán's Fidesz on 55 and the far-right Our Homeland on six.
For two years Magyar took his burgeoning movement around villages, town squares and cities, rallying Hungarians who had had enough of the cronyism and corruption that had become endemic over years.
"Never before in the history of democratic Hungary have so many people voted - and no single party has ever received such a strong mandate," he said on Sunday night, after a record 79% of the electorate turned out to vote.
Orbán's rule was built up through four successive election victories and sweeping majorities, but it was over in a matter of minutes.
As pro-Magyar supporters waited expectantly in the square on the Buda side of the Danube, the Tisza leader posted an extraordinary message on Facebook: "Viktor Orbán just called me on the phone and congratulated us on our victory."
There was barely time to digest what had just happened and only 30% of the vote had so far been counted.
But moments later Orbán himself appeared on a stage in a conference centre a mile down-river on the other side of the Danube, surrounded by his glum-looking Fidesz party colleagues.
"The result of the election is clear and painful," he told them, thanking the estimated 2.5 million Hungarians who stuck by him. "The days ahead of us are for us to heal our wounds."
Word began to spread among Magyar supporters, and at the nearby hotel that had become party HQ, Tisza activists hugged each other.
Magyar has promised to reverse Orbán-era changes to education and health, tackle corruption, restore the independence of the judiciary and kill off the widely loathed system of patronage known as NER that helped enrich party loyalists and squander state resources.
To make those changes to the constitution, he needed a two-thirds majority of 133 seats, and although the final results are not yet in, latest results suggest Tisza is on course for 138.
Another target in Magyar's sights is pro-Orbán state media. The M1 TV channel has until now slavishly toed the party line, along with what were once independent websites bought up over time by Fidesz allies.
Apparently uncertain what to do next, M1 rebroadcast a speech that Magyar had given just after the vote had finished. He had been hopeful of victory at the time, but by now the speech was out of date - he had already won.
Hungary has for some time felt like two different worlds running in tandem. In one, Orbán convinced his supporters and TV viewers they were heading for victory and four more years in power, backed up by opinion polls run by sympathetic pollsters, who continued to forecast a Fidesz victory as late as Sunday evening.
In the other world was Magyar, attracting big crowds wherever he went, backed up by respected pollsters reporting an increasing lead over his rival.
On Sunday night, those two worlds collided, and only one was real.
If Tisza's two-thirds majority is confirmed, state TV is set for change.
Magyar moved slowly through the crowds before taking to the stage, flanked by his party colleagues.
"You performed a miracle today, Hungary made history today," he told the audience, to chants of "Ria-Ria-Hungaria!"
He likened their electoral victory to the Hungarian revolution of 1848, and the uprising against Soviet occupation in 1956.
There was a time when Orbán himself spoke out against Soviet occupation, but he has become a close partner of Russia's Vladimir Putin, and his justification of cheap Russian oil and petrol has made him a highly unpopular leader in the EU, which has tried to shake off its reliance on Russian imports. Orbán has also reneged on an EU agreement to provide Ukraine with a €90bn (£79bn; $105bn) loan.
"Russians go home," Magyar supporters chanted, as the next Hungarian prime minister promised better relations with the European Union.
Poland's Prime Minister Donald Tusk was among the first of many European leaders to welcome Magyar's "glorious victory", adding in Hungarian "Ruszkik Haza" - Russians go home.
The Tisza leader pledged that his first trip abroad as prime minister would be to Warsaw - to reinforce Hungary's 1,000-year friendship with Poland.
As for Orbán, now 62, he has not resigned, and without him it is difficult to imagine what will happen to Fidesz.
For now he will continue to lead Hungary in a caretaker role, while he and his party lick their wounds.
▸ 展开全文
04-13 01:47 · AI应用,内容管理,社交媒体,用户工具,技术趋势
Bouncer: Block "crypto", "rage politics", and more from your X feed using AI
Heal your feed. Bouncer is a browser extension that uses AI to filter unwanted posts from your Twitter/X feed. Define filter topics in plain language — "crypto", "engagement bait", "rage politics" — and Bouncer classifies and hides matching posts in real time.
- Natural language filters — describe what you don't want to see in your own words
- Multiple AI backends — run models locally on your GPU, or use cloud APIs (OpenAI, Google Gemini, Anthropic, OpenRouter)
- On-device inference — local models via WebLLM run entirely in your browser with zero data sent externally
- Image-aware filtering — multimodal models can classify posts based on images, not just text
- Reasoning transparency — see exactly why each post was filtered
- Theme-aware UI — adapts to light, dim, and dark modes automatically
Local models are downloaded once and cached in the browser's Cache Storage.
Install Bouncer from the Chrome Web Store.
cd Bouncer
npm install
npm run build
- Open
chrome://extensions
- Enable Developer mode
- Click Load unpacked and select the
Bouncer/
folder - Navigate to twitter.com / x.com
- Click "Settings" in the Bouncer element and add your preferred provider API key (or enable local models) and select your preferred model from the dropdown.
- A MutationObserver watches the Twitter feed for new posts
- Post text, images, and metadata are extracted via the Twitter adapter
- Posts are queued and sent to the selected AI model for classification against your filter topics
- The model returns a category match and reasoning for each post
- Matching posts are hidden with a fade-out animation and added to your filtered posts list
- Click View filtered to review hidden posts and see why each was filtered
Results are cached so re-encountering a post doesn't require another inference call.
▸ 展开全文
04-13 01:47 · AI治理,平台合规,交付风险,声誉系统,自动化运维
We have a 99% email reputation, but Gmail disagrees
We have a 99% email reputation. Gmail disagrees.
- Written:
- on
Oooooh boy. Let’s get this out of the way first. Email sucks.
Now to the how and the why. We’re builders. We love making tools to help designers and developers live a little bit easier. We’re pretty good at it. Marketing, though? We do our best, but the truth is, we don’t like to bother people.
Like a lot of small software companies, we use SendGrid to deliver our emails. We try our best to follow email best practices. We even have a 99% reputation score in SendGrid. Gold star. A+ student.
Gmail, however, did not get the memo.
Right before we hit send on our announcement emails for our new Build Awesome Kickstarter campaign, we took a deeper look at some of our recent email sends. Things had gone quiet. Not bouncing. Not throwing errors. Just… disappearing into Gmail’s spam folder like a ‘possum slipping into a vent.
In our recent crash course, here’s what we’ve learned about Gmail deliverability: it runs its own reputation system that has absolutely nothing to do with anyone else’s opinion of you. If you don’t do certain things “correctly” (meaning Gmail’s own definition), you get marked as spam.
Now, there are definitely folks who will choose to mark some of what we send as spam. And for them, rightly so. We get that. But this is not that. We’ve entered a black hole for Gmail deliverability. And since 90% (literally) of our email list goes to Gmail addresses… the results aren’t pretty. It looks like this has been happening to us for a while. We’re a small company of just over 20 people, and can’t watch everything all the time. We’d rather be making you new icons. So some of you may have missed things we were genuinely excited to share. That’s a big bummer.
(And yes… there are companies out there that can likely help us with that. Most tend to be out of our price range. So we’ve been doing a lot of this on our own.)
But here’s the part that really gets us. At our CORE, our instinct is to only email folks when we actually have something fun to share. A big release, something we’re excited about, news worth your time. That’d probably be every couple of months, if that. Respectful. Low noise. How we want to be treated. Like, genuinely, if we could, we would only very occasionally send a big email blast to our customers.
Turns out, the email gods hate that. To keep a sending IP “warm” and maintain deliverability, you’re expected to send constantly. Like… all the time. Which means the system actively punishes companies for respecting their customers’ inboxes. It’s a genuine catch-22: send too many emails and your reputation drops from complaints. Send too few and it drops from inactivity. Try to do the right thing and you get penalized either way. And. It. Is. Frustrating.
We’re working to fix our issues by culling old addresses, slowing our sends down, and making sure all of our i’s are dotted and t’s are crossed. It’s not a fast fix.
So if you haven’t heard from us recently… or if you’ve heard TOO MUCH from us recently, that’s why. We’re working on it. And we’ve got a lot of good stuff to catch you up on. In the meantime, please help spread the word about Build Awesome. It’s a genuinely cool product, and we hope you’ll like it. At the very least, watch the video.
P.S. If you suspect you might’ve missed some emails from us, mind doing a quick favor? In your email client, search for from:hello@m.fontawesome.com in:spam
and click the little “Report Not Spam” button. You’re awesome.
▸ 展开全文
04-13 00:06 · 大模型发布,多模态竞争,技术迭代,内容生成,AI基准测试
OpenAI推出支持视频输入和输出的新一代多模态AI模型，在多个基准测试中表现优异
OpenAI今日正式发布了其新一代多模态AI模型，该模型在视频理解和生成能力方面实现了重大突破。
**技术特性**：
1. **视频输入支持**：模型能够直接处理视频片段，理解其中的场景、动作、物体和人物关系
2. **视频生成能力**：可根据文本描述生成高质量、连贯的视频内容，支持多种风格和分辨率
3. **多模态融合**：在文本、图像、音频、视频四种模态间实现无缝转换和理解
4. **基准测试表现**：在多个国际基准测试中表现优异，特别是在视频问答和内容生成任务上
**应用场景**：
- **内容创作**：为视频制作、广告、教育内容提供AI辅助
- **监控分析**：实时分析监控视频，识别异常行为和物体
- **娱乐产业**：为游戏、电影制作提供AI生成内容
- **教育培训**：创建交互式学习材料和模拟场景
**技术突破**：
- 采用新的Transformer架构，专门优化视频序列处理
- 训练数据包含数百万小时的标注视频内容
- 推理效率提升40%，相比前代模型减少计算资源消耗
- 支持实时视频处理，延迟低于100毫秒
**行业影响**：
OpenAI此举将进一步推动多模态AI技术的发展，特别是在视频内容理解和生成领域。竞争对手包括Google的VideoPoet、Meta的Make-A-Video等。分析认为，视频AI市场将在未来3年内达到千亿美元规模。
该模型现已通过API向开发者和企业客户开放，标准版支持720p视频处理，专业版支持4K视频和实时流处理。
▸ 展开全文
04-13 00:01 · 中东,局势,AlJazeera,中东
US Navy to enforce blockade, Trump says after condemning Iran for blocking waterway, complaining of no deal in Pakistan.
Trump says US to blockade ships crossing Strait of Hormuz
President Donald Trump says the United States Navy will begin blockading the Strait of Hormuz “immediately” after peace talks between the US and Iran in Pakistan ended without an agreement.
Trump, in a social media post on Sunday, accused Iran of “extortion” and said the US Navy would also hunt down and interdict ships in international waters that have paid Iran a toll to traverse the strait. He said US forces will also begin clearing mines that he says Iran placed in the strait.
Recommended Stories
list of 3 items- list 1 of 3Iran must not charge tolls in Strait of Hormuz, UN maritime chief says
- list 2 of 3Oil tankers exit Strait of Hormuz amid fragile US-Iran ceasefire
- list 3 of 3US-Iran ceasefire talks: What are the key sticking points?
“So, there you have it, the meeting went well, most points were agreed to, but the only point that really mattered, NUCLEAR, was not,” Trump said. “Effective immediately, the United States Navy, the Finest in the World, will begin the process of BLOCKADING any and all Ships trying to enter, or leave, the Strait of Hormuz.”
Iran has essentially taken control over the Strait of Hormuz, a vital chokepoint for the global energy market, since the US and Israel launched a war on Iran on February 28.
Iran’s Islamic Revolutionary Guard Corps (IRGC) warned on Sunday that civilian vessels can cross the strait in compliance with “specific regulations”, but that military vessels approaching the strait would be considered in violation of the ceasefire and “dealt with severely”.
Traffic through the narrow strait has slowed to a trickle, nearly paralysing about one-fifth of the world’s oil and liquefied natural gas shipments and sending shock waves through the global economy.
Trump’s comments have raised concerns about the status of the fragile two-week ceasefire.
“These two posts by the US president are very bombastic, very long, very detailed, and seem to be filled with a great deal of frustration,” Al Jazeera correspondent Zein Basravi said from Dubai in the United Arab Emirates.
“There are no doubt alarm bells ringing here in the GCC [Gulf Cooperation Council] for countries that were hoping this ceasefire would extend beyond those two weeks and become a consistent, long-term negotiated peace that could perhaps lead to different relationships forming in the region, but certainly to an end to the violence we see.”
Mediators have urged both parties to prioritise diplomacy and avoid a return to fighting.
“I urge that the ceasefire be extended and talks continue,” Oman’s Foreign Minister Badr Albusaidi said in a social media post on Sunday. “Success may require everyone to make painful concessions, but this is nothing as compared to the pain of failure and war.”
Iran has denied US claims that two of its warships recently passed through the strait for mine-clearing operations, warning that any military vessels seeking to do so would receive a “strong response”. Trump called Iran’s control over the waterway “world extortion” in his social media post and added that any Iranian forces who fire at US forces or “peaceful vessels” would be “BLOWN TO HELL”.
He added that the blockade would involve unspecified “other countries”, and he would not allow Iran to benefit from the closure of the strait. Trump also stated in his social media post that the US Navy would interdict “every vessel in International Waters that has paid a toll to Iran”.
“We’re not going to let Iran make money on selling oil to people that they like,” Trump told Fox News in an interview later on Sunday. “It’s going to be all or none, and that’s the way it is.”
He added that the United Kingdom and “a couple of other countries” are sending minesweeper vessels to aid the US effort, an assertion that has not been confirmed by the British government.
Democratic lawmakers in the US have criticised Trump’s comments.
“I don’t understand how blockading the strait is somehow going to push the Iranians into opening it,” Senator Mark Warner, a Democrat, said during an interview with CNN.
“I don’t see the connection there,” he added.
Iran has continued to send its own ships through the strait since the war began and has allowed a handful of vessels from other countries to pass through. Iranian officials have also discussed setting up a toll system after the fighting ends, under which users would pay Iran a fee to cross the strait.
“Iran knows that controlling the Strait of Hormuz is still their main leverage,” Al Jazeera correspondent Ali Hashem reported from the Iranian capital of Tehran. “The higher the tensions, the more prices surge, and this affects the global economy.”
▸ 展开全文
04-13 00:01 · 中东,局势,AlJazeera,中东
Trump said the US Navy would immediately start blockading the Strait of Hormuz after ceasefire talks in Pakistan fail.
The US military says it will begin blockading Iranian ports on Monday at 14:00 GMT, allowing ships not coming or going to Iran to pass through the Strait of Hormuz.
Iran’s Revolutionary Guard says the Strait of Hormuz is open for civilian vessels, but says that military ships “will be dealt with severely”.
▸ 展开全文
04-13 00:01 · 中东,局势,BBC,中东
Twenty-one hours was not enough to end 47 years of hostility between Iran and the US, writes the BBC's Lyse Doucet.
After Iran talks falter, the big question is what happens next?
Twenty-one hours was not enough to end 47 years of hostility between Iran and the US.
The historic high-level talks in Islamabad, during a pause in weeks of grievous war, were always unlikely to end any other way.
Calling this marathon negotiating session a failure belies the scale of the challenge in narrowing wide gaps on complex issues ranging from age-old suspicion about Iran's nuclear programme to new challenges this war has thrown up - most of all Iran's control of the strategic Strait of Hormuz, whose closure is causing economic shocks worldwide.
To do a deal, they also needed to overcome a deep chasm of distrust.
A day ago, it wasn't even certain the two sides would meet, and even more, sit down in the same room.
A longstanding political taboo was broken.
The urgent question now is: what happens next?
What happens to the contested two-week ceasefire which pulled the world back from US President Donald Trump's alarming threat to destroy a "whole civilisation" in Iran?
Would the US president be ready to send his negotiators back to the bargaining table?
We're hearing reports from sources here in Islamabad that some conversations have continued after US Vice-President JD Vance boarded his plane at sunrise, declaring the US delegation had made their "final and best offer".
Will the US now escalate or negotiate?
We still don't know enough about what happened behind tightly closed doors in a five-star hotel in leafy locked-down Islamabad during talks that went on long into the night.
There are still few details on the disputes and discussion between the two sides, assisted by Pakistani mediators, the calls to and from experts, advisers, and, according to Vance, "dozens" of calls to Trump himself.
The vice-president spoke of the "core goal" of the US during his brief dawn news conference.
"We need to see an affirmative commitment that [Iran] will not seek a nuclear weapon and they will not seek the tools that would enable them to quickly achieve a nuclear weapon," he said.
During the last round of talks in February, before military strikes were unleashed again, Iran had offered new concessions including the dilution of its 440kg stockpile of uranium enriched to 60% - dangerously close to weapons-grade.
But it still insists on its "right" to enrich and hasn't been willing to give up that stockpile, now said to be buried deep in the rubble after US and Israeli air strikes last year.
It's also refused repeated demands to open the Strait of Hormuz - to allow the free flow of vital traffic in oil, gas and other essential goods - in the absence of a new agreement.
Both the US and the Iranian delegations came to Islamabad emboldened by their belief that theirs was the winning side in this war.
And they engaged knowing that, if they failed, there was the option to keep fighting – whatever the spiralling pain for their own people and a world reeling from the cost of this conflagration.
There was also what Dr Sanam Vakil of Chatham House describes as a "limited psychological understanding of the adversary and what compromises are needed for a real deal".
Vance spoke of good news – "we've had a number of substantive negotiations" - and there was bad news: "We have not reached an agreement."
And he made it clear that was "bad news for Iran much more than the United States of America".
Iran's foreign ministry spokesman Esmail Baghaei criticised the US's "excessive demands and unlawful requests" in a post on X.
And its parliamentary speaker Mohammad Bagher Ghalibaf, who led Iran's negotiating team, wrote that "the opposing side ultimately failed to gain the trust of the Iranian delegation in this round of negotiations".
Iran is indicating it's ready to keep talking. Pakistan's Foreign Minister Ishaq Dar urged all sides to uphold the fragile ceasefire and said they would continue their efforts to encourage dialogue - sentiments being echoed in other concerned capitals.
If history provides any lessons, the last time Iran reached a nuclear deal with the US and other world powers in 2015, it took 18 months of breakthroughs and breakdowns.
Trump has made it clear he doesn't want to get bogged down in protracted negotiations. Vance previously warned that the US would not be receptive if Tehran tried to "play us".
Pakistani journalist Kamran Yousef - in a legion of journalists who pulled all-nighters to provide non-stop coverage with very few details - declared that this round was one of "no breakthrough but no breakdown either".
The world waits for a verdict.
▸ 展开全文
04-13 00:01 · 国际,新闻,BBC,国际
The Nigerian Air Force has not addressed reports it struck a village market while targeting militants.
Civilians feared killed after reports of air strike on Nigerian market
Civilians are feared to have died after reports that Nigerian military jets struck a village market in the north-east of the country.
The attack took place near the border between Yobe and Borno states while an aircraft was said to be engaged in an operation against Islamist militants in the area, according to the Reuters news agency and local reports.
The Nigerian Air Force said it carried out "mop-up" air strikes on "identified terrorist locations" within the Jilli axis of Borno State on Saturday, but has not confirmed that it struck the market nor confirmed casualties.
The area is at the centre of the long-running Boko Haram insurgency, which has forced more than two million people to flee their homes and killed thousands.
The Nigerian Air Force's statement said follow-on strikes on Saturday were "aimed at fleeing remnants and regrouping cells seeking to exploit the difficult terrain".
There are conflicting reports of the number of casualties as a result of the strike on the market.
Citing a councillor, Lawan Zanna Nur Geidam, in Yobe's Geidam district, as well as three residents and an official from an international humanitarian agency, news agency Reuters said at least 200 people had been killed.
"It's a very devastating incident," said Zanna Nur Geidam, who said the injured had been taken to hospitals in Yobe and Borno.
Nigerian outlets the Sun and Punch reported sources saying that 10 people had been killed, while "several" others had sustained injuries.
But newspaper the Daily Trust reported that an eyewitness said at least 56 people had died, with 14 others hospitalised.
According to the outlet, the weekly market usually attracts traders and buyers from surrounding communities.
In a statement on Facebook on Sunday, the Yobe State Emergency Management Agency said it had received "preliminary reports" of an incident at Jilli Market in the Gubio local government area, "which reportedly resulted in casualties affecting some marketers".
It added that emergency response mechanisms had been "immediately evacuated", with assessment teams deployed to the area, but added that "details regarding the nature of the incident and the number of casualties remain unverified".
The government of the neighbouring Yobe state said some of its residents who went to the market "were affected".
"Some people from Geidam LGA [local government area] bordering Gubio LGA in Borno state who went to the Jilli weekly market were affected," Brigadier General Dahiru Abdulsalam said on Facebook.
Ahmed Ali, a 43-year-old market trader, told Reuters he was injured in the blast.
"I became so scared and attempted to run away, but a friend dragged me and we all lay on the ground," he said.
Amnesty International's Nigeria branch condemned the strike, which it said had killed "over 100 people".
"Launching air raids is not a legitimate law enforcement method by anyone's standard," a post on X said.
"Such reckless use of deadly force is unlawful, outrageous and lays bare the Nigerian military's shocking disregard for the lives of those it supposedly exists to protect."
Nigeria's north‑east has seen repeated incidents in recent years in which military air operations against Islamist insurgents have mistakenly hit civilians, including in villages, camps for displaced people and markets.
▸ 展开全文