|
huggingface.co
|
HF Weights
|
Анализировать url
|
|
github.com
|
GitHub
|
Анализировать url
|
|
arxiv.org
|
Jeopardy! Benchmark
|
Анализировать url
|
|
arxiv.org
|
Specialized Benchmarks
|
Анализировать url
|
|
arxiv.org
|
Olympiad Performance
|
Анализировать url
|
|
arxiv.org
|
EAIRA Evaluation
|
Анализировать url
|
|
arxiv.org
|
Chart Understanding
|
Анализировать url
|
|
huggingface.co
|
<div class="flex-1 pr-4"><span class="font-display text-[10px] font-bold uppercase tracking-wider text-teal bg-green/10 border border-green/20 rounded-full px-2.5 py-0.5 inline-block mb-2">71B params</span><h3 class="font-display text-lg font-bold text-ink mb-1 group-hover:text-teal transition-colors">AstroSage-Llama-3.1-70B</h3><p class="text-xs text-ink/65 leading-relaxed">Flagship reasoning model. Specialized on two decades of physics and astronomy literature to execute multi-step scientific reasoning.</p></div><span class="w-10 h-10 rounded-full bg-green/10 text-green group-hover:bg-green group-hover:text-white flex items-center justify-center transition-all shrink-0"><svg class="w-4 h-4" fill="none" stroke="currentColor" viewbox="0 0 24 24"><path stroke-linecap="round" stroke-linejoin="round" stroke-width="2.5" d="M14 5l7 7m0 0l-7 7m7-7H3"></path></svg></span>
|
Анализировать url
|
|
huggingface.co
|
<div class="flex-1 pr-4"><span class="font-display text-[10px] font-bold uppercase tracking-wider text-teal bg-green/10 border border-green/20 rounded-full px-2.5 py-0.5 inline-block mb-2">8B params</span><h3 class="font-display text-lg font-bold text-ink mb-1 group-hover:text-teal transition-colors">AstroSage-Llama-3.1-8B</h3><p class="text-xs text-ink/65 leading-relaxed">Compact and efficient domain-specific assistant. Brings frontier-class in-domain accuracy to local consumer-grade hardware.</p></div><span class="w-10 h-10 rounded-full bg-green/10 text-green group-hover:bg-green group-hover:text-white flex items-center justify-center transition-all shrink-0"><svg class="w-4 h-4" fill="none" stroke="currentColor" viewbox="0 0 24 24"><path stroke-linecap="round" stroke-linejoin="round" stroke-width="2.5" d="M14 5l7 7m0 0l-7 7m7-7H3"></path></svg></span>
|
Анализировать url
|
|
huggingface.co
|
<div class="flex-1 pr-4"><span class="font-display text-[10px] font-bold uppercase tracking-wider text-teal bg-green/10 border border-green/20 rounded-full px-2.5 py-0.5 inline-block mb-2">7B to 70B</span><h3 class="font-display text-lg font-bold text-ink mb-1 group-hover:text-teal transition-colors">AstroLLaMA-2 / 3</h3><p class="text-xs text-ink/65 leading-relaxed">The original foundation line for astronomy LLMs that established specialized pre-training and alignment practices.</p></div><span class="w-10 h-10 rounded-full bg-green/10 text-green group-hover:bg-green group-hover:text-white flex items-center justify-center transition-all shrink-0"><svg class="w-4 h-4" fill="none" stroke="currentColor" viewbox="0 0 24 24"><path stroke-linecap="round" stroke-linejoin="round" stroke-width="2.5" d="M14 5l7 7m0 0l-7 7m7-7H3"></path></svg></span>
|
Анализировать url
|
|
arxiv.org
|
<div><div class="flex items-center gap-3 mb-1.5"><span class="text-[10px] font-display font-bold uppercase tracking-wider border rounded-full px-2.5 py-0.5 bg-amber-500/10 text-amber-400 border-amber-500/25">AstroMLab 1</span><span class="text-xs text-white/40 ml-2">2024<!-- --> · arXiv:<!-- -->2407.11194</span></div><h3 class="font-display text-xl font-bold text-white leading-snug group-hover:text-green-light transition-colors">AstroMLab 1: Who Wins Astronomy Jeopardy!?</h3><p class="text-sm text-white/60 mt-1.5">The first astronomy-specific LLM benchmark and a broad evaluation of proprietary and open models.</p><p class="text-xs text-green-light font-semibold mt-2">Derived from ARA&A review articles to measure high-level scientific reasoning.</p></div>
|
Анализировать url
|
|
arxiv.org
|
<div><div class="flex items-center gap-3 mb-1.5"><span class="text-[10px] font-display font-bold uppercase tracking-wider border rounded-full px-2.5 py-0.5 bg-blue-500/10 text-blue-400 border-blue-500/25">AstroMLab 2</span><span class="text-xs text-white/40 ml-2">2024<!-- --> · arXiv:<!-- -->2409.19750</span></div><h3 class="font-display text-xl font-bold text-white leading-snug group-hover:text-green-light transition-colors">AstroMLab 2: AstroLLaMA-2-70B Model and Benchmarking Specialised LLMs for Astronomy</h3><p class="text-sm text-white/60 mt-1.5">First rigorous benchmarking of specialized astronomy LLMs; the first 70B-parameter astronomy model.</p><p class="text-xs text-green-light font-semibold mt-2">Demonstrates that continual pre-training pays off at scale over naive specialization.</p></div>
|
Анализировать url
|
|
arxiv.org
|
<div><div class="flex items-center gap-3 mb-1.5"><span class="text-[10px] font-display font-bold uppercase tracking-wider border rounded-full px-2.5 py-0.5 bg-teal-500/10 text-teal-400 border-teal-500/25">AstroSage-8B</span><span class="text-xs text-white/40 ml-2">2024<!-- --> · arXiv:<!-- -->2411.09012</span></div><h3 class="font-display text-xl font-bold text-white leading-snug group-hover:text-green-light transition-colors">AstroMLab 3: Achieving GPT-4o Level Performance in Astronomy with a Specialized 8B-Parameter Large Language Model</h3><p class="text-sm text-white/60 mt-1.5">AstroSage-8B — a domain-specialized assistant trained on two decades of astronomy literature.</p><p class="text-xs text-green-light font-semibold mt-2">Domain alignment creates compact models matching commercial models in-domain.</p></div>
|
Анализировать url
|
|
arxiv.org
|
<div><div class="flex items-center gap-3 mb-1.5"><span class="text-[10px] font-display font-bold uppercase tracking-wider border rounded-full px-2.5 py-0.5 bg-purple-500/10 text-purple-400 border-purple-500/25">AstroSage-70B</span><span class="text-xs text-white/40 ml-2">2025<!-- --> · arXiv:<!-- -->2505.17592</span></div><h3 class="font-display text-xl font-bold text-white leading-snug group-hover:text-green-light transition-colors">AstroMLab 4: Benchmark-Topping Performance in Astronomy Q&A with a 70B-Parameter Domain-Specialized Reasoning Model</h3><p class="text-sm text-white/60 mt-1.5">AstroSage-70B — reasoning-capable, domain-specialized, evaluated against 119 models.</p><p class="text-xs text-green-light font-semibold mt-2">Combines domain expertise with reasoning traces for complex scientific Q&A.</p></div>
|
Анализировать url
|
|
arxiv.org
|
<div><div class="flex items-center gap-3 mb-1.5"><span class="text-[10px] font-display font-bold uppercase tracking-wider border rounded-full px-2.5 py-0.5 bg-emerald-500/10 text-emerald-400 border-emerald-500/25">AstroMLab 5</span><span class="text-xs text-white/40 ml-2">2025<!-- --> · arXiv:<!-- -->2511.12353</span></div><h3 class="font-display text-xl font-bold text-white leading-snug group-hover:text-green-light transition-colors">AstroMLab 5: Structured Summaries and Concept Extraction for 400,000 Astrophysics Papers</h3><p class="text-sm text-white/60 mt-1.5">A knowledge layer over all of astro-ph: structured summaries plus a curated concept vocabulary.</p><p class="text-xs text-green-light font-semibold mt-2">Bridges unstructured literature text and structured databases for AI search agents.</p></div>
|
Анализировать url
|
|
arxiv.org
|
<div><div class="text-[10px] text-green-light font-bold uppercase tracking-wider mb-1">arXiv:<!-- -->2602.14335</div><div class="font-display text-sm font-semibold text-white leading-snug mb-2">Predicting New Concept-Object Associations by Mining the Literature</div><p class="text-xs text-white/55 leading-relaxed">Forecasting concept-object knowledge graph links using ALS with similarity smoothing.</p></div><div class="text-[11px] text-white/40 mt-4 border-t border-white/5 pt-2">J. Li, Ting, Accomazzi, Ghosal, Ramachandra</div>
|
Анализировать url
|
|
arxiv.org
|
<div><div class="text-[10px] text-green-light font-bold uppercase tracking-wider mb-1">arXiv:<!-- -->2512.01270</div><div class="font-display text-sm font-semibold text-white leading-snug mb-2">Egent: An Autonomous Agent for Equivalent Width Measurement</div><p class="text-xs text-white/55 leading-relaxed">Autonomous spectral line fitting with LLM visual QA.</p></div><div class="text-[11px] text-white/40 mt-4 border-t border-white/5 pt-2">Ting, Saad, Liu, Shen</div>
|
Анализировать url
|
|
arxiv.org
|
<div><div class="text-[10px] text-green-light font-bold uppercase tracking-wider mb-1">arXiv:<!-- -->2510.08354</div><div class="font-display text-sm font-semibold text-white leading-snug mb-2">Mephisto: Automated Interpretation of Galaxy Observations</div><p class="text-xs text-white/55 leading-relaxed">Multi-agent tree search emulating human reasoning on SED fitting.</p></div><div class="text-[11px] text-white/40 mt-4 border-t border-white/5 pt-2">Sun, Ting, Liang, Duan, Huang, Cai</div>
|
Анализировать url
|
|
arxiv.org
|
<div><div class="text-[10px] text-green-light font-bold uppercase tracking-wider mb-1">arXiv:<!-- -->2510.05016</div><div class="font-display text-sm font-semibold text-white leading-snug mb-2">LLMs Achieve Gold Medal Performance at the IOAA</div><p class="text-xs text-white/55 leading-relaxed">Evaluating state-of-the-art models on astronomy Olympiad theory exams.</p></div><div class="text-[11px] text-white/40 mt-4 border-t border-white/5 pt-2">Pinheiro, Chen, Piazza, Shroff, Liang, Ting, Sun</div>
|
Анализировать url
|
|
arxiv.org
|
<div><div class="text-[10px] text-green-light font-bold uppercase tracking-wider mb-1">arXiv:<!-- -->2508.10075</div><div class="font-display text-sm font-semibold text-white leading-snug mb-2">Teaching LLMs to Speak Spectroscopy</div><p class="text-xs text-white/55 leading-relaxed">Efficiently repurposing LLaMA-3.1-8B to predict galaxy redshifts via LoRA.</p></div><div class="text-[11px] text-white/40 mt-4 border-t border-white/5 pt-2">Ramachandra, Ting, Sun, Wells, Habib</div>
|
Анализировать url
|
|
arxiv.org
|
<div><div class="text-[10px] text-green-light font-bold uppercase tracking-wider mb-1">arXiv:<!-- -->2506.06921</div><div class="font-display text-sm font-semibold text-white leading-snug mb-2">Teaching Astronomy with Large Language Models</div><p class="text-xs text-white/55 leading-relaxed">AI literacy integration in undergraduate education and AstroTutor.</p></div><div class="text-[11px] text-white/40 mt-4 border-t border-white/5 pt-2">Ting & O'Briain</div>
|
Анализировать url
|
|
arxiv.org
|
<div><div class="text-[10px] text-green-light font-bold uppercase tracking-wider mb-1">arXiv:<!-- -->2502.20309</div><div class="font-display text-sm font-semibold text-white leading-snug mb-2">EAIRA: Evaluating AI Models as Scientific Research Assistants</div><p class="text-xs text-white/55 leading-relaxed">Argonne-led multi-faceted evaluation framework for scientific assistants.</p></div><div class="text-[11px] text-white/40 mt-4 border-t border-white/5 pt-2">Cappello, Madireddy, Underwood, Ting, Wells, Foster, Stevens</div>
|
Анализировать url
|
|
arxiv.org
|
<div><div class="text-[10px] text-green-light font-bold uppercase tracking-wider mb-1">arXiv:<!-- -->2508.06492</div><div class="font-display text-sm font-semibold text-white leading-snug mb-2">Effective Training Data Synthesis for Improving MLLM Chart Understanding</div><p class="text-xs text-white/55 leading-relaxed">A 5-step data synthesis pipeline for multimodal chart reading.</p></div><div class="text-[11px] text-white/40 mt-4 border-t border-white/5 pt-2">Yang, Zhang, Hou, Li, Liu, Payani, Ting, Zheng</div>
|
Анализировать url
|
|
huggingface.co
|
Browse the models
|
Анализировать url
|
|
huggingface.co
|
Hugging Face
|
Анализировать url
|
|
huggingface.co
|
Benchmark dataset
|
Анализировать url
|
|
github.com
|
Knowledge graph
|
Анализировать url
|
|
github.com
|
Agents for Astronomy
|
Анализировать url
|
|
github.com
|
GitHub
|
Анализировать url
|
