
Tensorplex Labs
395 posts

Tensorplex Labs
@TensorplexLabs
Empowering humanity through decentralized intelligence — one epoch at a time. Products: https://t.co/1JhctOJYye Community: https://t.co/k0NBfNT5t3


The team has decided to suspend development of the Dojo subnet. Dojo was originally designed to decentralise the pipeline for high-quality human feedback and model fine-tuning. Throughout its development, we achieved several significant technical milestones, most notably the release of DOJO-INTERFACE-CODER-7B and our open-source contributions of human feedback data. These efforts were critical to our mission of narrowing the gap between open-source AI and proprietary, state-of-the-art models. Tackling one of the most complex problems in decentralised AI has provided us with invaluable lessons. We would like to share some factors that led to this decision: Whilst working with several startups on the monetisation of human feedback collection, we found that the technical cost of tailoring anti-gaming mechanisms to specific client needs made the current model economically unviable. It remains a significant challenge to build a framework that is simultaneously generalisable across modalities and rigorous enough for enterprise standards. We believe that for decentralisation to be meaningful, it must be built upon a foundation of tangible market utility. Consequently, we are shifting our immediate focus to building commercial pipelines and validating monetisation strategies. Tensorplex will be restructuring to focus our full attention on this "utility-first" strategy. Bittensor’s greatest strength is the diversity of its ecosystem. While our focus has shifted towards commercialisation, we believe the network thrives on a plurality of approaches - whether through research, monetisation, or attention. Our path is one of many attempts to drive the collective ecosystem forward. We hope that our insights, combined with our access to a broader enterprise network, will enable us to contribute even more effectively to the network in the future. The Dojo subnet repository will remain public as an open-source archive for anyone wishing to fork the code and continue the ambitious mission of decentralising the human interface for Bittensor. We’ll be more than happy to support independent teams with insights from our journey, helping to ensure that the progress made on Dojo remains a valuable resource for the entire Bittensor ecosystem. We want to extend our deepest gratitude to every miner, validator, developer, and community member who contributed their talents, resources, and feedback. Your passion is what built Dojo, and that spirit will continue to influence our future work. Onwards, Tensorplex Labs





Here are the Top 5 Bittensor Subnet gainers in the past 7 days! (06 Oct - 13 Oct) 1. SN121 @sundaebar_ai +188.73% 2. SN45 @Gen42AI +68.38% 3. SN46 @resilabsai +66.52% 4. SN60 @bitsecai +48.84% 5. SN66 @OceansSN66 +42.61% $tao













Bittensor Ecosystem Recap for the week! (Part 1/3) Upcoming Bittensor AI Event in Toronto - @opentensor Sportstensor Dashboard Upgrade Unveiled (Subnet 41) - @sportstensor @lcnxyz Syth Miners ouperform GMB benchmark by 25% - @SynthdataCo @modenetwork (Subnet 50)

Bittensor Ecosystem Recap for the week! (Part 1/3) Upcoming Bittensor AI Event in Toronto - @opentensor Sportstensor Dashboard Upgrade Unveiled (Subnet 41) - @sportstensor @lcnxyz Syth Miners ouperform GMB benchmark by 25% - @SynthdataCo @modenetwork (Subnet 50)


Hash Rate - Ep 124: Backprop / Tensorplex / Dojo (sn52) 🧙 Guest: CK @ckcheung of @TensorplexLabs @BackpropFinance $TAO 00:00 Introduction and Overview of Backprop 04:27 The Vision Behind Tensorplex and Backprop 10:14 Dojo (subnet 52): Bridging Human Taste and AI 17:32 Show and Tell: Visualizing Dojo's Functionality 27:05 Telegram Bot and Future Developments 27:17 Simplifying Crypto Trading with Telegram 30:21 Decentralized AI: A Bottom-Up Approach 39:06 Understanding Dojo's Role in Bittensor 44:12 Funding and Growth in the Bittensor Ecosystem 48:24 The Controversy of Burning Miner Emissions






