MAROKO133 Eksklusif ai: ChatGPT Is Saying VWeird Things in Chinese Edisi Jam 02:47

📌 MAROKO133 Breaking ai: ChatGPT Is Saying VWeird Things in Chinese Wajib Baca

If you thought English-language ChatGPT-prose was annoying, just wait until you get a load of its conversational habits in Chinese.

A fascinating piece of reporting by Wired took a look at how ChatGPT handles Chinese, the global language with the highest number of native speakers, according to the Language School at Middlebury College.

One of its go-to tics, the publication reports, is to answer questions with “我会稳稳地接住你,” which literally translates to “I will catch you steadily,” a phrase signalling willingness to talk about a person’s feelings (as Wired’s Zeyi Yang notes, a more flowerly translation could be “I’ll hold you steadily through whatever comes,” though the sentiment is apparently irritating to Chinese speakers either way.)

At other times, ChatGPT will tell its Chinese users “砍一刀,” which means either “help me cut it once,” or “slash the price,” an obnoxious bit of ad copy parroted by Chinese eCommerce platform Pinduoduo, Wired notes.

The odd mannerisms are so ubiquitous that they’ve become a meme among Chinese netizens, with some depicting ChatGPT as a giant inflatable airbag placed to break someone’s fall — catching them steadily.

The problem, Wired observes, may come down to a phenomenon called “mode collapse,” a fundamental bias tracing back to the people training large language models (LLMs). Basically, the idea goes that when human data annotators comb through text to train AI chatbots, they unknowingly favor familiar turns-of-phrase over more foreign-sounding sentences.

After an LLM like ChatGPT is trained, it becomes difficult to force it to “unlearn” certain phrases. While AI developers can reinforce the usage of a particular response — “I will catch you steadily” may be a great answer in a particular situation — accounting for range and quantity is a different beast altogether.

“We don’t know how to say: ‘this is good writing, but if we do this good writing thing 10 times, then it’s no longer good writing,” Max Spero, cofounder and CEO of AI-writing detector Pangram, told Wired.

Whatever the cause, it’s nice to know when English-speakers agree on something with our Chinese brethren, even if it’s just a universal hatred for ChatGPT’s inane babble.

More on ChatGPT: Even After Two Massacres, OpenAI Still Hasn’t Stopped ChatGPT From Helping Plan School Shootings

The post ChatGPT Is Saying VWeird Things in Chinese appeared first on Futurism.

🔗 Sumber: futurism.com

📌 MAROKO133 Hot ai: Adobe Research Unlocking Long-Term Memory in Video World Model

Video world models, which predict future frames conditioned on actions, hold immense promise for artificial intelligence, enabling agents to plan and reason in dynamic environments. Recent advancements, particularly with video diffusion models, have shown impressive capabilities in generating realistic future sequences. However, a significant bottleneck remains: maintaining long-term memory. Current models struggle to remember events and states from far in the past due to the high computational cost associated with processing extended sequences using traditional attention layers. This limits their ability to perform complex tasks requiring sustained understanding of a scene.

A new paper, “Long-Context State-Space Video World Models” by researchers from Stanford University, Princeton University, and Adobe Research, proposes an innovative solution to this challenge. They introduce a novel architecture that leverages State-Space Models (SSMs) to extend temporal memory without sacrificing computational efficiency.

The core problem lies in the quadratic computational complexity of attention mechanisms with respect to sequence length. As the video context grows, the resources required for attention layers explode, making long-term memory impractical for real-world applications. This means that after a certain number of frames, the model effectively “forgets” earlier events, hindering its performance on tasks that demand long-range coherence or reasoning over extended periods.

The authors’ key insight is to leverage the inherent strengths of State-Space Models (SSMs) for causal sequence modeling. Unlike previous attempts that retrofitted SSMs for non-causal vision tasks, this work fully exploits their advantages in processing sequences efficiently.

The proposed Long-Context State-Space Video World Model (LSSVWM) incorporates several crucial design choices:

1. Block-wise SSM Scanning Scheme: This is central to their design. Instead of processing the entire video sequence with a single SSM scan, they employ a block-wise scheme. This strategically trades off some spatial consistency (within a block) for significantly extended temporal memory. By breaking down the long sequence into manageable blocks, they can maintain a compressed “state” that carries information across blocks, effectively extending the model’s memory horizon.
2. Dense Local Attention: To compensate for the potential loss of spatial coherence introduced by the block-wise SSM scanning, the model incorporates dense local attention. This ensures that consecutive frames within and across blocks maintain strong relationships, preserving the fine-grained details and consistency necessary for realistic video generation. This dual approach of global (SSM) and local (attention) processing allows them to achieve both long-term memory and local fidelity.

The paper also introduces two key training strategies to further improve long-context performance:

• Diffusion Forcing: This technique encourages the model to generate frames conditioned on a prefix of the input, effectively forcing it to learn to maintain consistency over longer durations. By sometimes not sampling a prefix and keeping all tokens noised, the training becomes equivalent to diffusion forcing, which is highlighted as a special case of long-context training where the prefix length is zero. This pushes the model to generate coherent sequences even from minimal initial context.
• Frame Local Attention: For faster training and sampling, the authors implemented a “frame local attention” mechanism. This utilizes FlexAttention to achieve significant speedups compared to a fully causal mask. By grouping frames into chunks (e.g., chunks of 5 with a frame window size of 10), frames within a chunk maintain bidirectionality while also attending to frames in the previous chunk. This allows for an effective receptive field while optimizing computational load.

The researchers evaluated their LSSVWM on challenging datasets, including Memory Maze and Minecraft, which are specifically designed to test long-term memory capabilities through spatial retrieval and reasoning tasks.

The experiments demonstrate that their approach substantially surpasses baselines in preserving long-range memory. Qualitative results, as shown in supplementary figures (e.g., S1, S2, S3), illustrate that LSSVWM can generate more coherent and accurate sequences over extended periods compared to models relying solely on causal attention or even Mamba2 without frame local attention. For instance, on reasoning tasks for the maze dataset, their model maintains better consistency and accuracy over long horizons. Similarly, for retrieval tasks, LSSVWM shows improved ability to recall and utilize information from distant past frames. Crucially, these improvements are achieved while maintaining practical inference speeds, making the models suitable for interactive applications.

The Paper Long-Context State-Space Video World Models is on arXiv

The post Adobe Research Unlocking Long-Term Memory in Video World Models with State-Space Models first appeared on Synced.

🔗 Sumber: syncedreview.com