Fi
Fine-grained stylistic control of LLMs using model arithmetic
Hacker NewsFine-grained stylistic control of LLMs using model arithmetic
We developed a new framework that enables flexible control of generated text in language models. By combining several models and/or system prompts in one mathematical formula, it lets you tweak your style and combine model outputs with ease. A handy tool for those working with LLMs, looking for more fine-grained control of stylistic output. More details in our paper: https://arxiv.org/abs/2311.14479 . Feedback and potential applications are welcome.
Share cardActual performance
85points
8comments
Made the leaderboard
Launch Intel predictions
Analyze your own launch →
Indie HackersFits the IH revenue-focused audience · Missing: supports, reddit linkedin, podcasting86%86% predicted probability of success on Indie Hackers, based on ML models trained on real launch data.
best fitHighest predicted score across all platforms for this description.
Product HuntOn track for Day 1 leaderboard · Strong signals: model, new, models · Missing: mac, agents, macos86%86% predicted probability of success on Product Hunt, based on ML models trained on real launch data.
Hacker NewsMay not resonate with HN audience · Strong signals: io · Missing: https docs, excited, just released49%49% predicted probability of success on Hacker News, based on ML models trained on real launch data.
nativeThis product was originally launched on this platform.
TrustMRRLess likely to generate early MRR · Missing: mobile apps, ios, personal45%45% predicted probability of success on TrustMRR, based on ML models trained on real launch data.
AppSumoMay struggle as an AppSumo deal · Missing: plus, platform, intuitive43%43% predicted probability of success on AppSumo, based on ML models trained on real launch data.
Acquire.comPre-revenue stage for this audience · Missing: arr, mrr, revenue11%11% predicted probability of success on Acquire.com, based on ML models trained on real launch data.
BetaListMay not resonate with beta-testers · Missing: web3, chat, crypto3%3% predicted probability of success on BetaList, based on ML models trained on real launch data.
Incorrect prediction on native model
Similar products
Ps
Psychedelia fine tuned Stable Diffusion model63%Launch Intel prediction score: how likely this product is to succeed on its source platform, based on its name, tagline, and description.
Psychedelia fine tuned Stable Diffusion model
Hacker News5Us
Using LLMs to control robots63%Launch Intel prediction score: how likely this product is to succeed on its source platform, based on its name, tagline, and description.
Using LLMs to control robots
Hacker News2Mo
Model Validation Using LLMs45%Launch Intel prediction score: how likely this product is to succeed on its source platform, based on its name, tagline, and description.
Model Validation Using LLMs
Hacker News2Fi
Fine-grain resourcefulness40%Launch Intel prediction score: how likely this product is to succeed on its source platform, based on its name, tagline, and description.
Fine-grain resourcefulness
Hacker News3Fi
Fine Uploader40%Launch Intel prediction score: how likely this product is to succeed on its source platform, based on its name, tagline, and description.
Fine Uploader
Hacker News1Open Interface57%Launch Intel prediction score: how likely this product is to succeed on its source platform, based on its name, tagline, and description.
Control Your Computer Using LLMs
Indie Hackers1aiMo
Monitor LLMs Using LangChain and Infino47%Launch Intel prediction score: how likely this product is to succeed on its source platform, based on its name, tagline, and description.
Monitor LLMs Using LangChain and Infino
Hacker News1Ra
Raink – Document ranker using LLMs60%Launch Intel prediction score: how likely this product is to succeed on its source platform, based on its name, tagline, and description.
Raink – Document ranker using LLMs
Hacker News2Brassery56%Launch Intel prediction score: how likely this product is to succeed on its source platform, based on its name, tagline, and description.
Fine nameplates for fine art
Indie Hackers1$300/mosalesUs
Using LLMs and Embeddings to classify application errors80%Launch Intel prediction score: how likely this product is to succeed on its source platform, based on its name, tagline, and description.
Using LLMs and Embeddings to classify application errors
Hacker News65