Articles

Browse all AI articles. Use filters to narrow down results.

Showing 27201 to 27220 of 30029 entries.

Most Recent Articles

ArXiv NLP
cs.clresearcharxiv nlpllmraginference

Investigating Counterfactual Unfairness in LLMs towards Identities through Humor

By Shubin Kim, Yejin Son, Junyeong Park, Keummin Ka, Seungbeen Lee, Jaeyoung Lee, Hyeju Jang, Alice Oh, Youngjae Yu

arXiv:2604.18729v1 Announce Type: new Abstract: Humor holds up a mirror to social perception: what we find funny often reflects who we are and how we judge others. When language models engage with humor, their reactions expose the social assumptions they have internalized from training data. In this paper, we investigate counterfactual unfairness through humor by observing how the model's responses change when we swap who speaks and who is addressed while holding other factors constant. Our framework spans three tasks: humor generation refusal, speaker intention inference, and relational/societal impact prediction, covering both identity-agnostic humor and identity-specific disparagement humor. We introduce interpretable bias metrics that capture asymmetric patterns under identity swaps. Experiments across state-of-the-art models reveal consistent relational disparities: jokes told by privileged speakers are refused up to 67.5% more often, judged as malicious 64.7% more frequently, and rated up to 1.5 points higher in social harm on a 5-point scale. These patterns highlight how sensitivity and stereotyping coexist in generative models, complicating efforts toward fairness and cultural alignment.

ArXiv NLP
cs.clcs.airesearcharxiv nlpllminferencefine-tuning

Two-dimensional early exit optimisation of LLM inference

By Jan H\r{u}la, David Adamczyk, Tom\'a\v{s} Filip, Martin Pavl\'i\v{c}ek, Petr Sos\'ik

arXiv:2604.18592v1 Announce Type: new Abstract: We introduce a two-dimensional (2D) early exit strategy that coordinates layer-wise and sentence-wise exiting for classification tasks in large language models. By processing input incrementally sentence-by-sentence while progressively activating deeper layers, our method achieves multiplicative computational savings that exceed those from optimizing either dimension independently. Experimental evaluation across four state-of-the-art LLMs (Llama 3.1, Llama 3.2, Gemma, Qwen; 3B-8B parameters) on three sentiment classification datasets demonstrates additional speed-ups of 1.4--2.3$\times$ over optimal layer-wise early exit for simpler tasks with vanilla models, with graceful degradation on complex multi-class problems. Fine-tuning reduces but does not eliminate this advantage. The approach is model-agnostic, requires only lightweight classification adapters, and is orthogonal to complementary efficiency methods such as quantization and pruning. Our findings indicate that 2D early exit strategies excel when semantic information accumulates predictably across input structure, suggesting possible applicability to sequence-processing tasks beyond sentiment classification.

ArXiv NLP
cs.clresearcharxiv nlpmetanlp

Model-Agnostic Meta Learning for Class Imbalance Adaptation

By Hanshu Rao, Guangzeng Han, Xiaolei Huang

arXiv:2604.18759v1 Announce Type: new Abstract: Class imbalance is a widespread challenge in NLP tasks, significantly hindering robust performance across diverse domains and applications. We introduce Hardness-Aware Meta-Resample (HAMR), a unified framework that adaptively addresses both class imbalance and data difficulty. HAMR employs bi-level optimizations to dynamically estimate instance-level weights that prioritize genuinely challenging samples and minority classes, while a neighborhood-aware resampling mechanism amplifies training focus on hard examples and their semantically similar neighbors. We validate HAMR on six imbalanced datasets covering multiple tasks and spanning biomedical, disaster response, and sentiment domains. Experimental results show that HAMR achieves substantial improvements for minority classes and consistently outperforms strong baselines. Extensive ablation studies demonstrate that our proposed modules synergistically contribute to performance gains and highlight HAMR as a flexible and generalizable approach for class imbalance adaptation. Code is available at https://github.com/trust-nlp/ImbalanceLearning.

Product Hunt
startupproduct hunt

ConsoleMini

By moamen basel

Discussion | Link

Simon Willison Weblog
aisimon willison weblogclaudeanthropicllmragagentagents

<p><strong><a href="https://github.blog/news-insights/company-news/changes-to-github-copilot-individual-plans/">Changes to GitHub Copilot Individual plans</a></strong></p> On the same day as Claude Code's temporary will-they-won't-they $100/month kerfuffle (for the moment, <a href="https://simonwillison.net/2026/Apr/22/claude-code-confusion/#they-reversed-it">they won't</a>), here's the latest on GitHub Copilot pricing.</p> <p>Unlike Anthropic, GitHub put up an official announcement about their changes, which include tightening usage limits, pausing signups for individual plans (!), restricting Claude Opus 4.7 to the more expensive $39/month "Pro+" plan, and dropping the previous Opus models entirely.</p> <p>The key paragraph:</p> <blockquote> <p>Agentic workflows have fundamentally changed Copilot’s compute demands. Long-running, parallelized sessions now regularly consume far more resources than the original plan structure was built to support. As Copilot’s agentic capabilities have expanded rapidly, agents are doing more work, and more customers are hitting usage limits designed to maintain service reliability.</p> </blockquote> <p>It's easy to forget that just six months ago heavy LLM users were burning an order of magnitude less tokens. Coding agents consume a <em>lot</em> of compute.</p> <p>Copilot was also unique (I believe) among agents in charging per-request, not per-token. (<em>Correction: Windsurf also operated a credit system like this which they <a href="https://windsurf.com/blog/windsurf-pricing-plans">abandoned last month</a></em>.) This means that single agentic requests which burn more tokens cut directly into their margins. The most recent pricing scheme addresses that with token-based usage limits on a per-session and weekly basis.</p> <p>My one problem with this announcement is that it doesn't clearly clarify <em>which</em> product called "GitHub Copilot" is affected by these changes. Last month in <a href="https://teybannerman.com/strategy/2026/03/31/how-many-microsoft-copilot-are-there.html">How many products does Microsoft have named 'Copilot'? I mapped every one</a> Tey Bannerman identified 75 products that share the Copilot brand, 15 of which have "GitHub Copilot" in the title.</p> <p>Judging by the linked <a href="https://github.com/features/copilot/plans">GitHub Copilot plans page</a> this covers Copilot CLI, Copilot cloud agent and code review (features on <a href="https://github.com/">GitHub.com</a> itself), and the Copilot IDE features available in VS Code, Zed, JetBrains and more. <p><small></small>Via <a href="https://news.ycombinator.com/item?id=47838508">Hacker News</a></small></p> <p>Tags: <a href="https://simonwillison.net/tags/github">github</a>, <a href="https://simonwillison.net/tags/microsoft">microsoft</a>, <a href="https://simonwillison.net/tags/ai">ai</a>, <a href="https://simonwillison.net/tags/generative-ai">generative-ai</a>, <a href="https://simonwillison.net/tags/github-copilot">github-copilot</a>, <a href="https://simonwillison.net/tags/llms">llms</a>, <a href="https://simonwillison.net/tags/llm-pricing">llm-pricing</a>, <a href="https://simonwillison.net/tags/coding-agents">coding-agents</a></p>

Product Hunt
startupproduct huntagent

Agently

By Ben Lang

Discussion | Link

Product Hunt
startupproduct hunt

Blink AI CFO

By Garry Tan

Discussion | Link

Product Hunt
startupproduct huntclaude

Wrangle

By John Kreisher

Discussion | Link

Simon Willison Weblog
aisimon willison weblogopenaiclaudeanthropicgooglellmagentagents

<p>Anthropic today quietly (as in <em>silently</em>, no announcement anywhere at all) updated their <a href="https://claude.com/pricing">claude.com/pricing</a> page (but not their <a href="https://support.claude.com/en/articles/11049762-choosing-a-claude-plan">Choosing a Claude plan page</a>, which shows up first for me on Google) to add this tiny but significant detail (arrow is mine, <a href="https://simonwillison.net/2026/Apr/22/claude-code-confusion/#they-reversed-it">and it's already reverted</a>):</p> <p><img src="https://static.simonwillison.net/static/2026/anthropic-x.jpg" alt="Screenshot of the Claude pricing grid - Compare features across plans. Free, Pro, Max 5x and Max 20x all have the same features, with the exception of Claude Code which is on Max only and Claude Cowork which is on Pro and Max only. An arrow highlights the Claude Code for Pro cross." style="max-width: 100%;" /></p> <p>The <a href="https://web.archive.org/web/20260421040656/claude.com/pricing">Internet Archive copy</a> from yesterday shows a checkbox there. Claude Code used to be a feature of the $20/month Pro plan, but according to the new pricing page it is now exclusive to the $100/month or $200/month Max plans.</p> <p><em><strong>Update</strong>: don't miss <a href="https://simonwillison.net/2026/Apr/22/claude-code-confusion/#they-reversed-it">the update to this post</a>, they've already changed course a few hours after this change went live.</em></p> <p>So what the heck is going on? Unsurprisingly, <a href="https://www.reddit.com/r/ClaudeAI/comments/1srzhd7/psa_claude_pro_no_longer_lists_claude_code_as_an/">Reddit</a> and <a href="https://news.ycombinator.com/item?id=47854477">Hacker News</a> and <a href="https://twitter.com/i/trending/2046718768634589239">Twitter</a> all caught fire.</p> <p>I didn't believe the screenshots myself when I first saw them - aside from the pricing grid I could find no announcement from Anthropic anywhere. Then Amol Avasare, Anthropic's Head of Growth, <a href="https://twitter.com/TheAmolAvasare/status/2046724659039932830">tweeted</a>:</p> <blockquote> <p>For clarity, we're running a small test on ~2% of new prosumer signups. Existing Pro and Max subscribers aren't affected.</p> </blockquote> <p>And that appears to be the closest we have had to official messaging from Anthropic.</p> <p>I don't buy the "~2% of new prosumer signups" thing, since everyone I've talked to is seeing the new pricing grid and the Internet Archive has already <a href="https://web.archive.org/web/20260422001250/https://claude.com/pricing">snapped a copy</a>. Maybe he means that they'll only be running this version of the pricing grid for a limited time which somehow adds up to "2%" of signups?</p> <p>I'm also amused to see Claude Cowork remain available on the $20/month plan, because Claude Cowork is effectively a rebranded version of Claude Code wearing a less threatening hat!</p> <p>There are a whole bunch of things that are bad about this.</p> <p>If we assume this is indeed a test, and that test comes up negative and they decide not to go ahead with it, the damage has still been extensive:</p> <ol> <li>A whole lot of people got scared or angry or both that a service they relied on was about to be rug-pulled. There really is a significant difference between $20/month and $100/month for most people, especially outside of higher salary countries.</li> <li>The uncertainty is really bad! A tweet from an employee is <em>not</em> the way to make an announcement like this. I wasted a solid hour of my afternoon trying to figure out what had happened here. My trust in Anthropic's transparency around pricing - a <em>crucial factor</em> in how I understand their products - has been shaken.</li> <li>Strategically, should I be taking a bet on Claude Code if I know that they might 5x the minimum price of the product?</li> <li>More of a personal issue, but one I care deeply about myself: I invest a <a href="https://simonwillison.net/tags/claude-code/">great deal of effort</a> (that's 105 posts and counting) in teaching people how to use Claude Code. I don't want to invest that effort in a product that most people cannot afford to use.</li> </ol> <p>Last month I ran <a href="https://simonw.github.io/nicar-2026-coding-agents/">a tutorial for journalists</a> on "Coding agents for data analysis" at the annual NICAR data journalism conference. I'm not going to be teaching that audience a course that depends on a $100/month subscription!</p> <p>This also doesn't make sense to me as a strategy for Anthropic. Claude Code <em>defined the category</em> of coding agents. It's responsible for billions of dollars in annual revenue for Anthropic already. It has a stellar reputation, but I'm not convinced that reputation is strong enough for it to lose the $20/month trial and jump people directly to a $100/month subscription.</p> <p>OpenAI have been investing heavily in catching up to Claude Code with their Codex products. Anthropic just handed them this marketing opportunity on a plate - here's Codex engineering lead <a href="https://twitter.com/thsottiaux/status/2046740759056162816">Thibault Sottiaux</a>:</p> <blockquote> <p>I don't know what they are doing over there, but Codex will continue to be available both in the FREE and PLUS ($20) plans. We have the compute and efficient models to support it. For important changes, we will engage with the community well ahead of making them.</p> <p>Transparency and trust are two principles we will not break, even if it means momentarily earning less. A reminder that you vote with your subscription for the values you want to see in this world.</p> </blockquote> <p>I should note that I pay $200/month for Claude Max and I consider it well worth the money. I've had periods of free access in the past courtesy of Anthropic but I'm currently paying full price, and happy to do so.</p> <p>But I care about the accessibility of the tools that I work with and teach. If Codex has a free tier while Claude Code starts at $100/month I should obviously switch to Codex, because that way I can use the same tool as the people I want to teach how to use coding agents.</p> <p>Here's what I think happened. I think Anthropic are trying to optimize revenue growth - obviously - and someone pitched making Claude Code only available for Max and higher. That's clearly a bad idea, but "testing" culture says that it's worth putting even bad ideas out to test just in case they surprise you.</p> <p>So they started a test, without taking into account the wailing and gnashing of teeth that would result when their test was noticed - or accounting for the longer-term brand damage that would be caused.</p> <p>Or maybe they <em>did</em> account for that, and decided it was worth the risk.</p> <p>I don't think that calculation was worthwhile. They're going to have to make a <em>very</em> firm commitment along the lines of "we heard your feedback and we commit to keeping Claude Code available on our $20/month plan going forward" to regain my trust.</p> <p>As it stands, Codex is looking like a much safer bet for me to invest my time in learning and building educational materials around.</p> <h4 id="they-reversed-it">Update: they've reversed it already</h4> <p>In the time I was <em>typing this blog entry</em> Anthropic appear to have reversed course - the <a href="https://claude.com/pricing">claude.com/pricing page</a> now has a checkbox back in the Pro column for Claude Code. I can't find any official communication about it though.</p> <p>Let's see if they can come up with an explanation/apology that's convincing enough to offset the trust bonfire from this afternoon!</p> <h4 id="update-2">Update 2: it may still affect 2% of signups?</h4> <p>Amol <a href="https://x.com/TheAmolAvasare/status/2046788872517066971">on Twitter</a>:</p><blockquote><p>was a mistake that the logged-out landing page and docs were updated for this test [<a href="https://twitter.com/TheAmolAvasare/status/2046783926920978681">embedded self-tweet</a>]</p> <blockquote><p>Getting lots of questions on why the landing page / docs were updated if only 2% of new signups were affected.</p> <p>This was understandably confusing for the 98% of folks not part of the experiment, and we've reverted both the landing page and docs changes.</p> </blockquote> </blockquote> <p>So the experiment is still running, just not visible to the rest of the world?</p> <p>Tags: <a href="https://simonwillison.net/tags/ai">ai</a>, <a href="https://simonwillison.net/tags/generative-ai">generative-ai</a>, <a href="https://simonwillison.net/tags/llms">llms</a>, <a href="https://simonwillison.net/tags/anthropic">anthropic</a>, <a href="https://simonwillison.net/tags/llm-pricing">llm-pricing</a>, <a href="https://simonwillison.net/tags/ai-ethics">ai-ethics</a>, <a href="https://simonwillison.net/tags/coding-agents">coding-agents</a>, <a href="https://simonwillison.net/tags/claude-code">claude-code</a>, <a href="https://simonwillison.net/tags/codex-cli">codex-cli</a></p>

Product Hunt
startupproduct hunt

kimiflare

By Sina Meraji

Discussion | Link

Business Insider
newsbusiness insidervision

Millennium founder Izzy Englander. SALT Yiming Zhang, a top quant researcher, resigned from Jump Trading last week. Zhang is expected to get funding from hedge fund Millennium as an external manager, sources said. Jump quants typically must sit out a lengthy non-compete period as long as two years upon leaving. A top quant researcher who resigned from systematic trading firm Jump Trading Group last week is getting backing from hedge fund giant Millennium Management, Business Insider has learned. Yiming Zhang, one of Jump's most senior researchers and a 17-year veteran of the firm, is expected to start as an external manager for Millennium on an exclusive basis after his non-compete restrictions expire, people familiar with the matter said. Jump, like many secretive quant trading firms, is well-known for enforcing lengthy non-compete provisions, which can last two years even for non-executive employees. A Millennium spokesman declined to comment. Zhang did not respond to a request for comment. A Jump representative did not respond to a request for comment. After the resignations last week of Zhang and fellow senior researcher Darko Kirovski, Jump cofounder Paul Gurinas told Business Insider in a statement: "Yiming and Darko were instrumental in Jump's growth over the last 17 years, and we'll be forever grateful for that. They will be friends of the firm forever." Zhang and Kirovski were influential researchers at Jump Core Strategies, one of the trading firm's largest and most profitable units. Kirovski's plans after leaving Jump could not be learned. He did not respond to a request for comment. Millennium, which manages $84 billion, has hundreds of internal teams trading across the spectrum of hedge fund strategies. In recent years, however, it has also become one of the biggest backers of external hedge fund talent, seeding experienced investors with capital, often through separately managed accounts. Read the original article on Business Insider

Business Insider
newsbusiness insideropenaichatgptagentagents

OpenAI's first workspace, Greg Brockman's former residence at 2977 21st St., is on the market for $1.5 million. Courtesy Open House Photography OpenAI's first workspace, Greg Brockman's former residence, is on the market for $1.5 million. The unit at 2977 21st St. had its first open house and is attracting attention for its history. The residence features two floors, three bedrooms, two bathrooms, and two outdoor decks. You could own the birthplace of OpenAI — if you have $1.5 million to spare. When OpenAI first launched in 2015, its founders, including Sam Altman, did not have an office and primarily worked out of cofounder Greg Brockman's apartment in the Mission. That very loft apartment on the upper levels of 2977 21st St. in San Francisco has been listed for sale for $1,545,000 and had its first open house over the weekend. Shane Ray, the listing agent with Compass, told Business Insider that while some visitors arrive knowing the history of the unit, which is in a Victorian house built in 1885, others are simply attracted to the floor plan and are pleasantly surprised when they discover the association with one of the world's most valuable companies. "I think real estate is one part just logistics and a transaction, but there's a part of it for everybody that's a little bit emotional," said Ray. "I don't know if this changes the price of the property, but even if it's not that OpenAI started in that living room, any history that you can know about a place always adds a little bit to your understanding of the place and how you view the place," Ray added. The building 2977 21st Street is a Victorian house built in 1885 and renovated in 2015. Open Homes/Apple Photos Clean Up The upper unit spans two floors and occupies about two-thirds of the building's available space. It was fully renovated by the building's current owner in 2015. There's also a commercial unit and another apartment on the ground level, respectively occupied by an architecture firm and a longtime tenant. According to Ray, the building's owner is a photographer who bought the building more than a decade ago with the intention of making the upper unit his studio, but later changed his mind. Brockman became his first tenant and eventually left during the pandemic. The layout The 1,816-square-foot apartment features three bedrooms, two bathrooms, and two outdoor decks. Courtesy Open House Photography The 1,816-square-foot apartment features three bedrooms, two bathrooms, and two outdoor decks. The floor plan centers on a spacious open-concept main level, where a vaulted great room with exposed beams and wide-plank white oak flooring. The upper windows bring abundant natural light directly into the main floor. Upstairs, the primary suite has direct access to a deck overlooking a shared garden for all building residents. Two additional bedrooms are located at the rear of the main level, along with a second full bathroom. A look inside The unit has both Victorian and modern features. Courtesy Open House Photography Ray said the unit's interior is "a nice mix of old and new," with Victorian features like honeycomb windows, but a bright, open, modern interior with updated appliances. The interior features an in-unit washer and dryer, an open kitchen with an island, wooden beams, a bathtub, and built-in closets in the primary bedroom on the second floor. Pricing and availability An outdoor deck overlooks a shared garden and offers views of nature in San Francisco. Courtesy Open House Photography According to Ray, the owner has listed the entire building for a little under $2 million, with an option to purchase only the upper unit. There is no HOA for the building because all units are owned by a single individual, but if only the upper unit sells, there will be a plan to co-own the building. "We're likely setting an offer date later this week," said Ray. "Most of the people who are interested seem to be interested in the full building, and we've had a lot of interest." Read the original article on Business Insider

Techmeme
techtechmeme

Mark Gurman / Bloomberg: A look at Apple's management changes and several key executives' futures; sources say Mike Rockwell has considered leaving or taking an advisory role next year  —  John Ternus, Apple Inc.'s next chief executive officer, is up against a major challenge as he settles into the role: retaining top talent during a time of historic turnover.

Techmeme
techtechmeme

Financial Times: Top law firm Sullivan & Cromwell told a US federal bankruptcy court that a major filing it made in a high-profile case contained multiple AI hallucinations  —  Firm whose partners bill more than $2,000 per hour apologises to judge for software-driven errors in bankruptcy case

OpenAI Blog
researchaiopenai blogopenai

OpenAI Privacy Filter is an open-weight model for detecting and redacting personally identifiable information (PII) in text with state-of-the-art accuracy

Techmeme
techtechmeme

Tyler Kingkade / NBC News: The Los Angeles Unified School District's board votes to require screen time limits, making it the first major American school system to do so  —  The resolution followed pressure from parents who said their children developed unhealthy habits after schools required them to use iPads and laptops every day.

Techmeme
techtechmemeclaudeanthropic

Ed Zitron / Ed Zitron's Where's Your Ed At: An Anthropic support doc no longer lists Claude Code as part of its $20/month Pro plan; Anthropic says it is running a small test on ~2% of new prosumer signups  —  Executive Summary:  — Anthropic appears to have removed access to Claude Code for its $20-a-month “Pro” Plans.