Fix organization logo URL when HTML is embedded on profile

README.md

@@ -9,7 +9,7 @@ pinned: false
 ---
 
 <div align="center">
-<img src="https://huggingface.co/spaces/omida-ai/README/resolve/main/assets/omida-mark.png" alt="Omida AI" width="88" />
+<img src="https://huggingface.co/spaces/omida-ai/README/resolve/dacc5ce439b9a1691d601fd4eea8863bb6a063e6/assets/omida-mark.png" alt="Omida AI" width="88" />
 <h1>Omida AI</h1>
 <p><strong>Superintelligence in humanity’s corner.</strong></p>
 <p><a href="https://omida.ai">Website</a> &nbsp; / &nbsp; <a href="https://huggingface.co/omida-ai/omida-r1">Omida R1</a> &nbsp; / &nbsp; <a href="https://www.omida.ai/research">Research</a></p>

index.html

@@ -1,3 +1,3 @@
 <!doctype html><html lang="en"><head><meta charset="utf-8"><meta name="viewport" content="width=device-width,initial-scale=1"><title>Omida AI</title><style>
 :root{color-scheme:light dark;--bg:#fafafa;--fg:#141414;--muted:#636363;--line:#dcdcdc}*{box-sizing:border-box}body{margin:0;background:var(--bg);color:var(--fg);font-family:Arial,Helvetica,sans-serif;line-height:1.6}main{max-width:860px;margin:auto;padding:64px 28px}header{padding-bottom:40px;border-bottom:1px solid var(--line)}.mark{width:88px;height:88px;object-fit:contain;border-radius:16px}h1{font-size:clamp(40px,8vw,72px);letter-spacing:-.065em;line-height:1;margin:24px 0 18px}h2{font-size:28px;letter-spacing:-.035em;line-height:1.2;margin:40px 0 16px}.mission{font-size:20px;margin:0 0 20px}a{color:inherit;text-underline-offset:4px}nav{display:flex;flex-wrap:wrap;gap:24px;font-size:14px}p{max-width:74ch}section{padding:4px 0 24px;border-bottom:1px solid var(--line)}.links{display:grid;grid-template-columns:repeat(2,minmax(0,1fr));gap:16px;margin:24px 0}.link{display:block;border:1px solid var(--line);padding:18px;text-decoration:none}.link strong{display:block}.link span{display:block;color:var(--muted);font-size:14px;margin-top:8px}.note{font-size:14px;color:var(--muted)}footer{padding-top:28px;font-size:14px}@media(max-width:560px){main{padding:36px 20px}.links{grid-template-columns:1fr}}@media(prefers-color-scheme:dark){:root{--bg:#111;--fg:#f1f1f1;--muted:#b0b0b0;--line:#353535}}
-</style></head><body><main><header><img class="mark" src="assets/omida-mark.png" alt="Omida AI"><h1>Omida AI</h1><p class="mission">Superintelligence in humanity’s corner.</p><nav><a href="https://omida.ai" target="_blank" rel="noopener">Website</a><a href="https://huggingface.co/omida-ai/omida-r1" target="_blank" rel="noopener">Omida R1</a><a href="https://www.omida.ai/research" target="_blank" rel="noopener">Research</a></nav></header><section><h2>Intelligence that can act.</h2><p>Omida is an independent AI research lab working toward systems that understand the world, reason over uncertainty, and carry complex technical work through to completion. Our focus connects world modeling and simulation with reasoning, planning, and reliable tool use.</p><p>This is the home of our public R-line: open-weight research models and the data, evaluation methods, and engineering behind them. The research track informs the development of future proprietary systems.</p></section><section><h2>Start with Omida R1</h2><p>Omida R1 is our first open-weight research release for operator-style technical work: investigating software, using tools, repairing code, and grounding conclusions in evidence.</p><div class="links"><a class="link" href="https://huggingface.co/omida-ai/omida-r1" target="_blank" rel="noopener"><strong>Model &amp; weights ↗</strong><span>The v0.1 Qwen3.5-9B fine-tune, lineage, license, and inference guidance.</span></a><a class="link" href="https://huggingface.co/omida-ai/omida-r1/blob/main/research/README.md" target="_blank" rel="noopener"><strong>Research package ↗</strong><span>Corpus generation, verification, evaluation, and a tool-use harness.</span></a><a class="link" href="https://huggingface.co/omida-ai/omida-r1/blob/main/docs/CORPUS.md" target="_blank" rel="noopener"><strong>Training corpus ↗</strong><span>Versioned operator demonstrations, provenance, schemas, and exports.</span></a><a class="link" href="https://huggingface.co/omida-ai/omida-r1/blob/main/docs/BENCHMARK.md" target="_blank" rel="noopener"><strong>Benchmark ↗</strong><span>Executable technical scenarios and outcome-verification methods.</span></a><a class="link" href="https://huggingface.co/omida-ai/omida-r1/blob/main/docs/REPRODUCTION.md" target="_blank" rel="noopener"><strong>Reproduce &amp; contribute ↗</strong><span>Reproduction instructions for the published research package.</span></a></div><p class="note">R1 is an early research checkpoint. The published weights are a fine-tune of Qwen3.5-9B; the newer corpus and benchmark are separate artifacts. Broad capability gains have not yet been established by published model evaluations.</p></section><section><h2>Research directions</h2><p><strong>World models &amp; simulation</strong><br>Representing systems, testing assumptions, and exploring how decisions change outcomes.</p><p><strong>Reasoning &amp; planning</strong><br>Solving unfamiliar problems and sustaining coherent work across longer tasks.</p><p><strong>Technical operators</strong><br>Software investigation, debugging, secure engineering, and authorized security research.</p><p><strong>Evaluation &amp; human agency</strong><br>Verifiable outcomes, calibrated uncertainty, and meaningful human oversight.</p></section><footer><strong>Explore the work. Run the experiments. Build with us.</strong><p><a href="https://omida.ai" target="_blank" rel="noopener">omida.ai</a> · <a href="https://huggingface.co/omida-ai/omida-r1/discussions" target="_blank" rel="noopener">Research discussions</a> · <a href="https://huggingface.co/omida-ai/omida-r1/blob/main/docs/CONTRIBUTING.md" target="_blank" rel="noopener">Contribute</a></p></footer></main></body></html>
+</style></head><body><main><header><img class="mark" src="https://huggingface.co/spaces/omida-ai/README/resolve/dacc5ce439b9a1691d601fd4eea8863bb6a063e6/assets/omida-mark.png" alt="Omida AI"><h1>Omida AI</h1><p class="mission">Superintelligence in humanity’s corner.</p><nav><a href="https://omida.ai" target="_blank" rel="noopener">Website</a><a href="https://huggingface.co/omida-ai/omida-r1" target="_blank" rel="noopener">Omida R1</a><a href="https://www.omida.ai/research" target="_blank" rel="noopener">Research</a></nav></header><section><h2>Intelligence that can act.</h2><p>Omida is an independent AI research lab working toward systems that understand the world, reason over uncertainty, and carry complex technical work through to completion. Our focus connects world modeling and simulation with reasoning, planning, and reliable tool use.</p><p>This is the home of our public R-line: open-weight research models and the data, evaluation methods, and engineering behind them. The research track informs the development of future proprietary systems.</p></section><section><h2>Start with Omida R1</h2><p>Omida R1 is our first open-weight research release for operator-style technical work: investigating software, using tools, repairing code, and grounding conclusions in evidence.</p><div class="links"><a class="link" href="https://huggingface.co/omida-ai/omida-r1" target="_blank" rel="noopener"><strong>Model &amp; weights ↗</strong><span>The v0.1 Qwen3.5-9B fine-tune, lineage, license, and inference guidance.</span></a><a class="link" href="https://huggingface.co/omida-ai/omida-r1/blob/main/research/README.md" target="_blank" rel="noopener"><strong>Research package ↗</strong><span>Corpus generation, verification, evaluation, and a tool-use harness.</span></a><a class="link" href="https://huggingface.co/omida-ai/omida-r1/blob/main/docs/CORPUS.md" target="_blank" rel="noopener"><strong>Training corpus ↗</strong><span>Versioned operator demonstrations, provenance, schemas, and exports.</span></a><a class="link" href="https://huggingface.co/omida-ai/omida-r1/blob/main/docs/BENCHMARK.md" target="_blank" rel="noopener"><strong>Benchmark ↗</strong><span>Executable technical scenarios and outcome-verification methods.</span></a><a class="link" href="https://huggingface.co/omida-ai/omida-r1/blob/main/docs/REPRODUCTION.md" target="_blank" rel="noopener"><strong>Reproduce &amp; contribute ↗</strong><span>Reproduction instructions for the published research package.</span></a></div><p class="note">R1 is an early research checkpoint. The published weights are a fine-tune of Qwen3.5-9B; the newer corpus and benchmark are separate artifacts. Broad capability gains have not yet been established by published model evaluations.</p></section><section><h2>Research directions</h2><p><strong>World models &amp; simulation</strong><br>Representing systems, testing assumptions, and exploring how decisions change outcomes.</p><p><strong>Reasoning &amp; planning</strong><br>Solving unfamiliar problems and sustaining coherent work across longer tasks.</p><p><strong>Technical operators</strong><br>Software investigation, debugging, secure engineering, and authorized security research.</p><p><strong>Evaluation &amp; human agency</strong><br>Verifiable outcomes, calibrated uncertainty, and meaningful human oversight.</p></section><footer><strong>Explore the work. Run the experiments. Build with us.</strong><p><a href="https://omida.ai" target="_blank" rel="noopener">omida.ai</a> · <a href="https://huggingface.co/omida-ai/omida-r1/discussions" target="_blank" rel="noopener">Research discussions</a> · <a href="https://huggingface.co/omida-ai/omida-r1/blob/main/docs/CONTRIBUTING.md" target="_blank" rel="noopener">Contribute</a></p></footer></main></body></html>