#LLM

Eight entries on this site carry the LLM tag: three papers and five posts, dated 2026. Explore the full list of related work below.

Papers

From Errors to Proofs: Minimal-Core-Guided Repair for Neuro-Symbolic Constraint Solving
Paper · 2026 · LogiSymb @ IJCAI-ECAI 2026 · cited by 0 When an LLM-generated constraint program is unsatisfiable, returning a minimal unsatisfiable core instead of an error message localises the fault and cuts fabricated solutions from 79% to 7%.
Test-Input Generation for Tensor Programs: What Actually Finds Kernel Bugs
Paper · 2026 · arXiv · cited by 0 Seven test-input generation strategies evaluated on the 26-op gpuemu corpus, reporting bug recall and control false-positive rate.
The Correctness Illusion in LLM-Generated GPU Kernels
Paper · 2026 · arXiv · cited by 6 Fixed-shape allclose checks used by LLM-kernel benchmarks certify buggy kernels as correct. Under op-schema-aware seeded fuzzing with an fp64 reference and per-(op, dtype) tolerances, a 26-op corpus shows 10 of 10 seeded illusions caught and 16 of 16 controls.

Why GPU Kernel Benchmarks Miss Real Bugs
Post · 2026 Passing a correctness benchmark and being correct in production are different claims for LLM-generated GPU kernels — the gap comes down to which inputs get sampled.
From AutoPrompt to TextGrad: A Prompt Survey
Post · 2026 A chronological survey of automated prompt optimization 2020–2025: AutoPrompt, APE, OPRO, EvoPrompt, DSPy, TextGrad, PromptAgent, and how to choose between them.
LLM Prompt Compression: A Practitioner Guide
Post · 2026 A practitioner's guide to LLM prompt compression: LLMLingua, GIST Tokens, 500xCompressor, KV-cache methods, and the rate-distortion limits of compressing context.
Prompt Structuring Techniques for LLMs
Post · 2026 A chronological survey of LLM prompt structuring: chain-of-thought, the instruction hierarchy, system prompt design, and evaluation frameworks.
Comparing LLM Safety Techniques
Post · 2026 A practitioner's survey of LLM safety techniques across OpenAI Harmony, Anthropic Constitutional AI, Google SAIF, Meta Llama Guard, and open-source RLHF frameworks.