SOTAVerified

Instruction Following

Instruction following is the basic task of the model. This task is dedicated to evaluating the ability of the large model to follow human instructions. It is hoped that the model can generate controllable and safe answers.

Papers

Recently AddedMost HypedMost ActiveNeeds VerificationMost Verified

Showing 101–125 of 1135 papers

TitleStatusHype
When Thinking Fails: The Pitfalls of Reasoning for Instruction-Following in LLMs—0
GuideBench: Benchmarking Domain-Oriented Guideline Following for LLM Agents—0
Navigating the Alpha Jungle: An LLM-Powered MCTS Framework for Formulaic Factor Mining—0
BLEUBERI: BLEU is a surprisingly effective reward for instruction followingCode1
MergeBench: A Benchmark for Merging Domain-Specialized LLMsCode1
UniEval: Unified Holistic Evaluation for Unified Multimodal Understanding and Generation—0
Tests as Prompt: A Test-Driven-Development Benchmark for LLM Code Generation—0
HealthBench: Evaluating Large Language Models Towards Improved Human HealthCode7
Judging the Judges: Can Large Vision-Language Models Fairly Evaluate Chart Comprehension and Reasoning?Code0
A Multi-Dimensional Constraint Framework for Evaluating and Improving Instruction Following in Large Language ModelsCode1
Efficient Telecom Specific LLM: TSLAM-Mini with QLoRA and Digital Twin Data—0
MM-Skin: Enhancing Dermatology Vision-Language Model with an Image-Text Dataset Derived from TextbooksCode1
Assessing Robustness to Spurious Correlations in Post-Training Language Models—0
Adaptive Markup Language Generation for Contextually-Grounded Visual Document UnderstandingCode1
T2VTextBench: A Human Evaluation Benchmark for Textual Control in Video Generation Models—0
LLaMA-Omni2: LLM-based Real-time Spoken Chatbot with Autoregressive Streaming Speech SynthesisCode3
Incentivizing Inclusive Contributions in Model Sharing Markets—0
PIPA: A Unified Evaluation Protocol for Diagnosing Interactive Planning Agents—0
T2VPhysBench: A First-Principles Benchmark for Physical Consistency in Text-to-Video Generation—0
UAV-VLN: End-to-End Vision Language guided Navigation for UAVs—0
Ask, Fail, Repeat: Meeseeks, an Iterative Feedback Benchmark for LLMs' Multi-turn Instruction-Following Ability—0
CachePrune: Neural-Based Attribution Defense Against Indirect Prompt Injection Attacks—0
TF1-EN-3M: Three Million Synthetic Moral Fables for Training Small, Open Language ModelsCode0
Breaking the Modality Barrier: Universal Embedding Learning with Multimodal LLMs—0
ParamΔ for Direct Weight Mixing: Post-Train Large Language Model at Zero Cost—0
Show:102550
← PrevPage 5 of 46Next →

Benchmark Results

#ModelMetricClaimedVerifiedStatus
1AutoIF (Llama3 70B)Inst-level loose-accuracy90.4—Unverified
2AutoIF (Qwen2 72B)Inst-level loose-accuracy88—Unverified
3GPT-4Inst-level loose-accuracy85.37—Unverified
4PaLM 2 SInst-level loose-accuracy59.11—Unverified