Zeroshot translation with LLM of Homer align more with Lattimore's translation, apparently a canonical one in the English speaking world? #chr2025 doi.org/10.63744/Xcj...
December 10, 2025 at 11:02 AM
Zeroshot translation with LLM of Homer align more with Lattimore's translation, apparently a canonical one in the English speaking world? #chr2025 doi.org/10.63744/Xcj...
zeroshot by deepseek v4.1f being told "also make the UI prettier" lol
September 16, 2026 at 11:33 AM
zeroshot by deepseek v4.1f being told "also make the UI prettier" lol
And for those curious this is a totally out of domain zeroshot example 🔥
August 2, 2023 at 8:52 PM
And for those curious this is a totally out of domain zeroshot example 🔥
I'm really enjoying the structured output ability of Gemini for this zero shot classification problem. I'm trying to categorize bioRxiv papers to focus on the ones most relevant to my interests and it's working with very little fiddling.
December 1, 2024 at 11:01 PM
I'm really enjoying the structured output ability of Gemini for this zero shot classification problem. I'm trying to categorize bioRxiv papers to focus on the ones most relevant to my interests and it's working with very little fiddling.
We realized this 6 months ago when we built zeroshot
June 22, 2026 at 2:42 PM
We realized this 6 months ago when we built zeroshot
Interested in LLMs for scalable zeroshot text&image annotation &analysis? The Baltic #DH summerschool published recordings, mine here: www.youtube.com/watch?v=Fm7m...
- Intro to LLMs
- quantitizing analytics framework
- assessing error rates
- using OpenAI APIs
- running your open LLM in free Colab
- Intro to LLMs
- quantitizing analytics framework
- assessing error rates
- using OpenAI APIs
- running your open LLM in free Colab
August 9, 2024 at 10:22 AM
Interested in LLMs for scalable zeroshot text&image annotation &analysis? The Baltic #DH summerschool published recordings, mine here: www.youtube.com/watch?v=Fm7m...
- Intro to LLMs
- quantitizing analytics framework
- assessing error rates
- using OpenAI APIs
- running your open LLM in free Colab
- Intro to LLMs
- quantitizing analytics framework
- assessing error rates
- using OpenAI APIs
- running your open LLM in free Colab
Robot Air Hockey Player Predicts Your Next Move
Robot Air Hockey Player Predicts Your Next Move
Air hockey is a fun game, but it’s one you can’t play by yourself. That is, unless you have a smart robot hockey player to act as your rival. [Zeroshot] …read more
hackaday.com
December 17, 2024 at 7:30 AM
Robot Air Hockey Player Predicts Your Next Move
🚀 Releasing a new zeroshot-classifier based on ModernBERT! Some key takeaways:
- ⚡ Speed & efficiency: It's multiple times faster and uses significantly less memory than DeBERTav3. You can use larger batch sizes and enabling bf16 (instead of fp16) gave me a ~2x speed boost
- 📉 Performance tradeoff:
- ⚡ Speed & efficiency: It's multiple times faster and uses significantly less memory than DeBERTav3. You can use larger batch sizes and enabling bf16 (instead of fp16) gave me a ~2x speed boost
- 📉 Performance tradeoff:
January 6, 2025 at 4:40 PM
🚀 Releasing a new zeroshot-classifier based on ModernBERT! Some key takeaways:
- ⚡ Speed & efficiency: It's multiple times faster and uses significantly less memory than DeBERTav3. You can use larger batch sizes and enabling bf16 (instead of fp16) gave me a ~2x speed boost
- 📉 Performance tradeoff:
- ⚡ Speed & efficiency: It's multiple times faster and uses significantly less memory than DeBERTav3. You can use larger batch sizes and enabling bf16 (instead of fp16) gave me a ~2x speed boost
- 📉 Performance tradeoff:
similar note on the laziness (RL rew hacking result I presume), they really need to be told NOT to take shortcuts everywhere (similar to specifying detailed acceptance tests up front): wan will just create slop if asked for zeroshot pleo brainrot, but it can do a decent job of bg visuals, talking
March 13, 2026 at 3:38 PM
similar note on the laziness (RL rew hacking result I presume), they really need to be told NOT to take shortcuts everywhere (similar to specifying detailed acceptance tests up front): wan will just create slop if asked for zeroshot pleo brainrot, but it can do a decent job of bg visuals, talking
We collaborated with a private sector media house to explore quantifying media balance&polarization using #ML #LLMs, on the example of stance towards immigration. Part of this was done pre-chatgpt, so we present results comparing finetuned BERTlikes vs zeroshot GPTs
journals.plos.org/plosone/arti...
journals.plos.org/plosone/arti...
April 30, 2024 at 8:41 AM
We collaborated with a private sector media house to explore quantifying media balance&polarization using #ML #LLMs, on the example of stance towards immigration. Part of this was done pre-chatgpt, so we present results comparing finetuned BERTlikes vs zeroshot GPTs
journals.plos.org/plosone/arti...
journals.plos.org/plosone/arti...
AI agents crash when using 20 tools at once.
Others fix: LangChain Enterprise ($$$)
ZeroShot fix: Multi-agent orchestration with 14 parallel tools + controlled concurrency. No thrash. No crashes.
🔗 github.com/anix-lynch/ai-agent-job-intelligence-phase-2
#GenAI #Agents
Others fix: LangChain Enterprise ($$$)
ZeroShot fix: Multi-agent orchestration with 14 parallel tools + controlled concurrency. No thrash. No crashes.
🔗 github.com/anix-lynch/ai-agent-job-intelligence-phase-2
#GenAI #Agents
March 4, 2026 at 3:38 AM
AI agents crash when using 20 tools at once.
Others fix: LangChain Enterprise ($$$)
ZeroShot fix: Multi-agent orchestration with 14 parallel tools + controlled concurrency. No thrash. No crashes.
🔗 github.com/anix-lynch/ai-agent-job-intelligence-phase-2
#GenAI #Agents
Others fix: LangChain Enterprise ($$$)
ZeroShot fix: Multi-agent orchestration with 14 parallel tools + controlled concurrency. No thrash. No crashes.
🔗 github.com/anix-lynch/ai-agent-job-intelligence-phase-2
#GenAI #Agents
Our pick of the week by @mgaido91.bsky.social: "AlignFormer: Modality Matching Can Achieve Better Zero-shot Instruction-Following Speech-LLM" by Ruchao Fan, Bo Ren, Yuxuan Hu, Rui Zhao, Shujie Liu, Jinyu Li (2024).
#NLProc #Speech #instructionfollowing #zeroshot #speechtech #speechllm
#NLProc #Speech #instructionfollowing #zeroshot #speechtech #speechllm
very interesting to see more and more methods to close the length mismatch between speech and text sequences (aka length adapter -- see arxiv.org/abs/2402.12025) for SFM+LLM models! This one merging CTC and Q-former sounds very cool to me:
arxiv.org/abs/2412.01145
arxiv.org/abs/2412.01145
AlignFormer: Modality Matching Can Achieve Better Zero-shot Instruction-Following Speech-LLM
Integrating speech into LLM (speech-LLM) has gaining increased attention recently. The mainstream solution is to connect a well-trained speech encoder and LLM with a neural adapter. However, the lengt...
arxiv.org
February 14, 2025 at 10:38 AM
Our pick of the week by @mgaido91.bsky.social: "AlignFormer: Modality Matching Can Achieve Better Zero-shot Instruction-Following Speech-LLM" by Ruchao Fan, Bo Ren, Yuxuan Hu, Rui Zhao, Shujie Liu, Jinyu Li (2024).
#NLProc #Speech #instructionfollowing #zeroshot #speechtech #speechllm
#NLProc #Speech #instructionfollowing #zeroshot #speechtech #speechllm
Oh interesting! If not a secret, what's the approach you're evaluating? More like the BERTopic type vec clustering, or cluster+zeroshot (maartengr.github.io/BERTopic/get...), or text classification w preset topics? (we argue for the latter in arxiv.org/abs/2309.14379 & www.nature.com/articles/s41...
A framework for the analysis of historical newsreels - Humanities and Social Sciences Communications
Humanities and Social Sciences Communications - A framework for the analysis of historical newsreels
www.nature.com
May 29, 2024 at 11:18 PM
Oh interesting! If not a secret, what's the approach you're evaluating? More like the BERTopic type vec clustering, or cluster+zeroshot (maartengr.github.io/BERTopic/get...), or text classification w preset topics? (we argue for the latter in arxiv.org/abs/2309.14379 & www.nature.com/articles/s41...
Experience this piece for yourselves. Separately from the writing, I want to highlight two ways the creators make their processes transparent, something far too rare in news stories.
1) Explanation of data analysis, including how they used AI for topic analysis and for image analysis.
1) Explanation of data analysis, including how they used AI for topic analysis and for image analysis.
November 16, 2025 at 7:54 PM
Experience this piece for yourselves. Separately from the writing, I want to highlight two ways the creators make their processes transparent, something far too rare in news stories.
1) Explanation of data analysis, including how they used AI for topic analysis and for image analysis.
1) Explanation of data analysis, including how they used AI for topic analysis and for image analysis.
Simply one of the best #AI follows for technical #LLM and #MachineLearning available.
#DiscoverAI
#AgenticRAG
#ZeroShot #AgenticAI #Agentic
#ArXiv #ArXivCS Paper Guides
youtube.com/post/Ugkx2bj...
#DiscoverAI
#AgenticRAG
#ZeroShot #AgenticAI #Agentic
#ArXiv #ArXivCS Paper Guides
youtube.com/post/Ugkx2bj...
Post from Discover AI - YouTube
Hi community, latest paper on AgenticRAG w/ TOOLS (Oct 3rd). Tool-Augmented Foundation Models for Zero-Shot Explainable Recommender Systems. Authors: " Agent...
youtube.com
October 9, 2025 at 6:32 AM
Simply one of the best #AI follows for technical #LLM and #MachineLearning available.
#DiscoverAI
#AgenticRAG
#ZeroShot #AgenticAI #Agentic
#ArXiv #ArXivCS Paper Guides
youtube.com/post/Ugkx2bj...
#DiscoverAI
#AgenticRAG
#ZeroShot #AgenticAI #Agentic
#ArXiv #ArXivCS Paper Guides
youtube.com/post/Ugkx2bj...
Take your flight simulator setup to new heights with zeroshot's ATmega32U4-controlled force feedback joystick.
I built a next-level flight joystick - using Force Feedback
YouTube video by zeroshot
youtu.be
April 19, 2025 at 7:43 PM
Take your flight simulator setup to new heights with zeroshot's ATmega32U4-controlled force feedback joystick.
AI agents quietly get dumber after updates.
Others fix: LangSmith evals ($$$), Phoenix (paid)
ZeroShot fix: Regression checks around FastAPI serving. Ensemble ML (XGBoost, LightGBM, CatBoost) with time-aware validation.
🔗 github.com/anix-lynch/churn-ml-pipeline
#MLOps #GenAI
Others fix: LangSmith evals ($$$), Phoenix (paid)
ZeroShot fix: Regression checks around FastAPI serving. Ensemble ML (XGBoost, LightGBM, CatBoost) with time-aware validation.
🔗 github.com/anix-lynch/churn-ml-pipeline
#MLOps #GenAI
March 5, 2026 at 5:45 PM
AI agents quietly get dumber after updates.
Others fix: LangSmith evals ($$$), Phoenix (paid)
ZeroShot fix: Regression checks around FastAPI serving. Ensemble ML (XGBoost, LightGBM, CatBoost) with time-aware validation.
🔗 github.com/anix-lynch/churn-ml-pipeline
#MLOps #GenAI
Others fix: LangSmith evals ($$$), Phoenix (paid)
ZeroShot fix: Regression checks around FastAPI serving. Ensemble ML (XGBoost, LightGBM, CatBoost) with time-aware validation.
🔗 github.com/anix-lynch/churn-ml-pipeline
#MLOps #GenAI
if you're interested in events annotated with country-level dyadic interactions, we have an earlier paper focusing on doing that in a zeroshot setting for user-specified event types: arxiv.org/abs/2305.15051
A Monte Carlo Language Model Pipeline for Zero-Shot Sociopolitical Event Extraction
Current social science efforts automatically populate event databases of "who did what to whom?" tuples, by applying event extraction (EE) to text such as news. The event databases are used to analyze...
arxiv.org
July 7, 2025 at 4:56 PM
if you're interested in events annotated with country-level dyadic interactions, we have an earlier paper focusing on doing that in a zeroshot setting for user-specified event types: arxiv.org/abs/2305.15051
AI agents crash when using 20 tools at once.
Others fix: LangChain Enterprise ($$$)
ZeroShot fix: Multi-agent orchestration with 14 parallel tools + controlled concurrency. No thrash. No crashes.
🔗 github.com/anix-lynch/ai-agent-job-intelligence-phase-2
#GenAI #Agents
Others fix: LangChain Enterprise ($$$)
ZeroShot fix: Multi-agent orchestration with 14 parallel tools + controlled concurrency. No thrash. No crashes.
🔗 github.com/anix-lynch/ai-agent-job-intelligence-phase-2
#GenAI #Agents
March 4, 2026 at 5:17 PM
AI agents crash when using 20 tools at once.
Others fix: LangChain Enterprise ($$$)
ZeroShot fix: Multi-agent orchestration with 14 parallel tools + controlled concurrency. No thrash. No crashes.
🔗 github.com/anix-lynch/ai-agent-job-intelligence-phase-2
#GenAI #Agents
Others fix: LangChain Enterprise ($$$)
ZeroShot fix: Multi-agent orchestration with 14 parallel tools + controlled concurrency. No thrash. No crashes.
🔗 github.com/anix-lynch/ai-agent-job-intelligence-phase-2
#GenAI #Agents
AI agents crash when using 20 tools at once.
Others fix: LangChain Enterprise ($$$)
ZeroShot fix: Multi-agent orchestration with 14 parallel tools + controlled concurrency. No thrash. No crashes.
🔗 github.com/anix-lynch/ai-agent-job-intelligence-phase-2
#GenAI #Agents
Others fix: LangChain Enterprise ($$$)
ZeroShot fix: Multi-agent orchestration with 14 parallel tools + controlled concurrency. No thrash. No crashes.
🔗 github.com/anix-lynch/ai-agent-job-intelligence-phase-2
#GenAI #Agents
March 4, 2026 at 3:41 AM
AI agents crash when using 20 tools at once.
Others fix: LangChain Enterprise ($$$)
ZeroShot fix: Multi-agent orchestration with 14 parallel tools + controlled concurrency. No thrash. No crashes.
🔗 github.com/anix-lynch/ai-agent-job-intelligence-phase-2
#GenAI #Agents
Others fix: LangChain Enterprise ($$$)
ZeroShot fix: Multi-agent orchestration with 14 parallel tools + controlled concurrency. No thrash. No crashes.
🔗 github.com/anix-lynch/ai-agent-job-intelligence-phase-2
#GenAI #Agents
Our pick of the week by @beomseok-lee.bsky.social: "AlignFormer: Modality Matching Can Achieve Better Zero-shot Instruction-Following Speech-LLM" by
Ruchao Fan, Bo Ren, Yuxuan Hu, Rui Zhao, Shujie Liu, and Jinyu Li (2024).
#speech #LLM #speechlmm #zeroshot #instructionfollowing
Ruchao Fan, Bo Ren, Yuxuan Hu, Rui Zhao, Shujie Liu, and Jinyu Li (2024).
#speech #LLM #speechlmm #zeroshot #instructionfollowing
Should speech come before the instruction text, or should the instruction text come first in a speech-language model?
Find out the best positioning for speech and text—and the novel adapter that aligns speech and text modalities!
arxiv.org/abs/2412.01145
Find out the best positioning for speech and text—and the novel adapter that aligns speech and text modalities!
arxiv.org/abs/2412.01145
AlignFormer: Modality Matching Can Achieve Better Zero-shot Instruction-Following Speech-LLM
Integrating speech into LLM (speech-LLM) has gaining increased attention recently. The mainstream solution is to connect a well-trained speech encoder and LLM with a neural adapter. However, the lengt...
arxiv.org
April 3, 2025 at 10:46 AM
Our pick of the week by @beomseok-lee.bsky.social: "AlignFormer: Modality Matching Can Achieve Better Zero-shot Instruction-Following Speech-LLM" by
Ruchao Fan, Bo Ren, Yuxuan Hu, Rui Zhao, Shujie Liu, and Jinyu Li (2024).
#speech #LLM #speechlmm #zeroshot #instructionfollowing
Ruchao Fan, Bo Ren, Yuxuan Hu, Rui Zhao, Shujie Liu, and Jinyu Li (2024).
#speech #LLM #speechlmm #zeroshot #instructionfollowing
🚨 New benchmark update! Latest AI models O3, O4-mini & Gemini 2.5 Flash were tested on classic games like Sokoban, Phoenix Wright, Candy Crush & 2048.
🔥 O3 led the zero-shot leaderboard, beating past top models!
#AI #MachineLearning #ZeroShot #GameAI #Benchmark
🔥 O3 led the zero-shot leaderboard, beating past top models!
#AI #MachineLearning #ZeroShot #GameAI #Benchmark
April 25, 2025 at 11:08 AM
🚨 New benchmark update! Latest AI models O3, O4-mini & Gemini 2.5 Flash were tested on classic games like Sokoban, Phoenix Wright, Candy Crush & 2048.
🔥 O3 led the zero-shot leaderboard, beating past top models!
#AI #MachineLearning #ZeroShot #GameAI #Benchmark
🔥 O3 led the zero-shot leaderboard, beating past top models!
#AI #MachineLearning #ZeroShot #GameAI #Benchmark
🤖 #Zeroshot creates autonomous engineering teams for #ClaudeCode #AI #DevTools #opensource
🔧 Point at a #GitHub issue, walk away, and return to verified, production-ready PRs
🛡️ Multiple isolated agents check each other's work - validators didn't write the code, so they can't lie about tests
🧵 👇
🔧 Point at a #GitHub issue, walk away, and return to verified, production-ready PRs
🛡️ Multiple isolated agents check each other's work - validators didn't write the code, so they can't lie about tests
🧵 👇
January 13, 2026 at 12:45 AM
🤖 #Zeroshot creates autonomous engineering teams for #ClaudeCode #AI #DevTools #opensource
🔧 Point at a #GitHub issue, walk away, and return to verified, production-ready PRs
🛡️ Multiple isolated agents check each other's work - validators didn't write the code, so they can't lie about tests
🧵 👇
🔧 Point at a #GitHub issue, walk away, and return to verified, production-ready PRs
🛡️ Multiple isolated agents check each other's work - validators didn't write the code, so they can't lie about tests
🧵 👇