[AINews] AI gets Memory • ButtondownTwitterTwitter

The section discusses summaries of Discord channels related to AI topics. It covers various discussions such as exploring new language models, roleplay performances, finetuning techniques, JavaScript and Python integration, model training losses, RAM limitations, medical LLMs, GPU optimizations, future model support, beta releases, multi-model management, reproduction of Magvit V2 model, AI-generated imagery legal issues, deepfake pornography, AI image standards, checksums for data shards, image content classification tools, collaborative paper review, cloud computing resources, and research computing collaboration opportunities.

Perplexity AI Outshines Rivals in Complex Query Handling

• Perplexity AI Outshines Rivals in Complex Query Handling: @tbrams tested Perplexity AI with a difficult question from the 'Gemini' paper and found it outperformed Google's Gemini service and OpenAI, answering more quickly. The test results from Perplexity AI are documented here.
• Perplexity's Potential in API Customization Highlighted: The PPLX API allows for custom search queries using parameters like 'site:reddit.com OR site:youtube.com', as mentioned by @me.lk. However, several users have encountered issues with the API such as performance hiccups (@andrewgazelka) and nonsensical responses (@myadmingushwork_52332).
• Perplexity AI Subscription and Renewal Queries Addressed: Users are seeking details on trial subscriptions and renewal processes for Pro subscriptions, with inquiries about token refresh rates also surfacing. There is currently no early access program for new Perplexity features as confirmed by @icelavaman.
• Promising Enhancements and Community Collaborations: Perplexity AI is receiving community praise for tools like the pplx shortcut action (@twodogseeds). Meanwhile, @ok.alex is encouraging a community-driven effort to contribute to an alternative feed/newsletter Alt-D-Feed.
• Seeking Direct Support Channel for Sensitive Data Issues: A user (@kitsuiwebster) has expressed the need for direct assistance with a sensitive company data issue, avoiding public disclosure while lacking response from support channels.

Reka Enters the Model Arena:

A new AI entity named the Reka model has sparked interest in the community following a tweet shared by @swyxio. The excitement is palpable with discussions around the tweet found here.

Investor Insights Meet AI:

@swyxio spotlighted a VC podcast delving into AI, which could be of significant interest to engineering aficionados. The podcast episode is accessible here.

BUD-E Buzz:

BUD-E, an empathetic and context-aware open voice assistant developed by LAION, could signal a new direction in conversational AI. More details are laid out on the LAION blog.

Pondering the Definition of Agents:

The community exchanged views on defining 'agents,' with @slono suggesting that they are goal-oriented programs that require minimal input from users, a concept significant in the realm of AI development.

Karpathy's OpenAI Exit Raises Questions:

The AI community is abuzz over the news of Andrej Karpathy leaving OpenAI, with @nembal pointing to an article from The Information and speculation about AGI influences. The article is accessible here.

Understanding Finetuning Techniques:

• @starsupernova explained that Mixtral – Instruct was trained using SFT on an instruction dataset followed by Direct Preference Optimization (DPO) on a paired feedback dataset, as detailed on page 6 of their paper. DPO is described as an optimized form of RLHF/PPO finetuning.

In the channel 'interpretability-general,' user @jaimerv asked for a more current overview of approaches to interpretability.

This section explores various discussions happening in different Discord channels related to topics such as contributions to leaderboards, concerns about data alignment in Pythia, updates on the LlamaIndex framework, building RAG applications, and recent announcements and innovations by HuggingFace. Participants are engaging in conversations about specific tasks, collaboration opportunities, performance optimizations, and new product releases within the AI community.

Hierarchical Image Classification Challenge:

• User @cropinky discussed the complexity of hierarchical image classification and emphasized the importance of data quality and quantity, suggesting further research on an ECCV22 paper and related datasets.

In Search of Gaussian Splats:

• User @aeros93 inquired about resources for creating Gaussian splats from point clouds or images, with user @johko990 suggesting redirecting the query to a more suitable channel for assistance.

Quest for Multimodal Project Insights:

• User @joee2711 sought clarification on Q-former / MLP connector differences and expressed interest in connecting with others working on similar multimodal projects.

Enhancing Image Retrieval Systems:

• User @femiloye is developing an image retrieval system akin to person reidentification and is seeking methods to enhance match accuracy beyond model embeddings, currently utilizing a custom deit transformer trained with reid loss.

DAMO-NLP-SG Releases Vast Long-Context Dataset:

• A dataset called LongCorpus-2.5B, containing 2.5B tokens collected from various domains for long-context continual pre-training, was shared.
• The dataset was inspired by Long-Data-Collections and ensures low n-gram similarity with the training set.

Scaling Models with 'rope' vs 'self-extend':

• 'Self-extend' was highlighted as preserving coherence better than 'rope scaling' even at larger scaling factors.

Ease of 'self-extend' Implementation:

• The benefits of 'self-extend' include no need for setup, fine-tuning, or extra parameters compared to 'gguf configurations' for quants.

• ethux pointed to an instance of Mixtral deployed on HuggingFace.co/chat, offering an alternate route to those facing cloud service barriers.
• Links mentioned:
  • Deploy with SkyPilot | Mistral AI Large Language Models: SkyPilot is a framework for running LLMs, AI, and batch jobs on any cloud, offering maximum cost savings, highest GPU availability, and managed execution.
  • casperhansen/mixtral-instruct-awq · Hugging Face: no description found
  • TheBloke/Mixtral-8x7B-Instruct-v0.1-AWQ · Hugging Face: no description found
  • HuggingChat: Making the community's best AI chat models available to everyone.
  • TheBloke/Mixtral-8x7B-Instruct-v0.1-AWQ · always getting 0 in output: no description found

This section discusses various topics related to prompt engineering and API infrastructure in the OpenAI Discord channels. It includes conversations about newbie guidance in prompt engineering, library queries for prompt engineering related to software development, crafting lightweight text adventures, generating jokes, and more. Additionally, it covers discussions on MPS support, Yi-34b training, model adaptation, LLM endpoint services, challenges with LLMs, and more in the OpenAccess AI Collective channels. Moreover, the challenges, enhancements, and proposed schema for MessagesList format in the axolotl channels are explored. Lastly, the LangChain AI section introduces a journaling app with memory and shares contributions on building a RAG application with NextJS, OpenAI API, and Dewy.

This section includes various discussions from different Discord channels related to topics like BM25 search method, CUDA experiments, AI models, and hackathon invitations. Users share experiences with BM25, implementations for CUDA on AMD GPUs, and discussions on matrix transposition for faster computation. Additionally, the section highlights news like Apple Silicon's performance monitoring tool, Andrej Karpathy's departure from OpenAI, and the addition of memory features in ChatGPT. Various links to GitHub repositories, articles, and tweets are also shared throughout the discussions.

This section contains links to the newsletter and social media channels related to AI News. It also mentions that the newsletter is brought to you by Buttondown, a platform for starting and growing newsletters.