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 91–100 of 1135 papers

TitleStatusHype
Meta SecAlign: A Secure Foundation LLM Against Prompt Injection AttacksCode2
VerIF: Verification Engineering for Reinforcement Learning in Instruction FollowingCode2
FusionAudio-1.2M: Towards Fine-grained Audio Captioning with Multimodal Contextual FusionCode2
When Large Multimodal Models Confront Evolving Knowledge:Challenges and PathwaysCode2
How Instruction and Reasoning Data shape Post-Training: Data Quality through the Lens of Layer-wise GradientsCode2
MM-IFEngine: Towards Multimodal Instruction FollowingCode2
Beyond Single-Turn: A Survey on Multi-Turn Interactions with Large Language ModelsCode2
CrystalFormer-RL: Reinforcement Fine-Tuning for Materials DesignCode2
Rec-R1: Bridging Generative Large Language Models and User-Centric Recommendation Systems via Reinforcement LearningCode2
LLaVA-MORE: A Comparative Study of LLMs and Visual Backbones for Enhanced Visual Instruction TuningCode2
Show:102550
← PrevPage 10 of 114Next →

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