LLM-AutoDiff: Auto-Differentiate Any LLM Workflow · HackerPapers discussions

LLM-AutoDiff: Auto-Differentiate Any LLM Workflow — mentioned by Xmd5a

Sun, 20 Jul 2025 11:22:34 GMT

>A database query might return 10,000 rows, but the agent only needs to know "query succeeded, 10k results, here are the first 5." Designing these abstractions is an art. It seems the author never used prompt/workflow optimization techniques. LLM-AutoDiff: Auto-Differentiate Any LLM Workflow https://arxiv.org/pdf/2501.16673

Paper and discussion history. A mention is not an endorsement.

LLM-AutoDiff: Auto-Differentiate Any LLM Workflow — mentioned by Xmd5a

Wed, 04 Jun 2025 11:20:27 GMT

This is exactly what I'm doing. Some papers I'm studying: TextGrad: Automatic "Differentiation" via Text: https://arxiv.org/abs/2406.07496 LLM-AutoDiff: Auto-Differentiate Any LLM Workflow : https://arxiv.org/abs/2501.16673 Trace is the Next AutoDiff: Generative Optimization with Rich Feedback, Execution Traces, and LLMs: https://arxiv.org/abs/2406.16218 GReaTer: Gradients over Reasoning Makes Smaller Language Models Strong Prompt Optimizers: https://arxiv.org/abs/2412.09722 PromptWizard: Task-Aware Prompt Optimization Framework: https://arxiv.org/abs/2405.18369

Paper and discussion history. A mention is not an endorsement.

LLM-AutoDiff: Auto-Differentiate Any LLM Workflow — mentioned by Xmd5a

Tue, 22 Apr 2025 13:59:14 GMT

Stumbled upon this in another thread: https://ai-analytics.wharton.upenn.edu/generative-ai-labs/re... My point still holds that it is optimizable though (https://github.com/zou-group/textgrad, https://arxiv.org/abs/2501.16673) >Subjects develop elaborate "rain dances" in the belief that they can influence the outcome. Not unlike sports fans superstitions. Anybody tuning neural weights by hand would feel like this.

Paper and discussion history. A mention is not an endorsement.

LLM-AutoDiff: Auto-Differentiate Any LLM Workflow — mentioned by meame2010

Wed, 05 Feb 2025 20:05:34 GMT

not few shot, but prompt tuning via text generation via auto-differentiation. https://arxiv.org/abs/2501.16673

Paper and discussion history. A mention is not an endorsement.

LLM-AutoDiff: Auto-Differentiate Any LLM Workflow — HN submission

Wed, 29 Jan 2025 05:15:45 GMT

Auto-Differentiating Any LLM Workflow: A Farewell to Manual Prompting

137 points; 32 discussion replies on HN. Paper and source history. A submission is not an endorsement.