SOTAVerified

Text Generation

Text Generation is the task of generating text with the goal of appearing indistinguishable to human-written text. This task is more formally known as "natural language generation" in the literature.

Text generation can be addressed with Markov processes or deep generative models like LSTMs. Recently, some of the most advanced methods for text generation include BART, GPT and other GAN-based approaches. Text generation systems are evaluated either through human ratings or automatic evaluation metrics like METEOR, ROUGE, and BLEU.

Further readings:

( Image credit: Adversarial Ranking for Language Generation )

Papers

Recently AddedMost HypedMost ActiveNeeds VerificationMost Verified

Showing 1–50 of 5335 papers

TitleStatusHype
Making Language Model a Hierarchical Classifier and GeneratorCode0
Mitigating Object Hallucinations via Sentence-Level Early InterventionCode1
The Devil behind the mask: An emergent safety vulnerability of Diffusion LLMsCode2
Seq vs Seq: An Open Suite of Paired Encoders and DecodersCode2
Hashed Watermark as a Filter: Defeating Forging and Overwriting Attacks in Weight-based Neural Network WatermarkingCode0
Exploiting Leaderboards for Large-Scale Distribution of Malicious Models—0
FIFA: Unified Faithfulness Evaluation Framework for Text-to-Video and Video-to-Text Generation—0
CLI-RAG: A Retrieval-Augmented Framework for Clinically Structured and Context Aware Text Generation with LLMs—0
Advancing Offline Handwritten Text Recognition: A Systematic Review of Data Augmentation and Generation Techniques—0
OpenFActScore: Open-Source Atomic Evaluation of Factuality in Text GenerationCode0
CoSteer: Collaborative Decoding-Time Personalization via Local Delta Steering—0
The Future is Agentic: Definitions, Perspectives, and Open Challenges of Multi-Agent Recommender Systems—0
CAVALRY-V: A Large-Scale Generator Framework for Adversarial Attacks on Video MLLMs—0
Compressed and Smooth Latent Space for Text Diffusion Modeling—0
Cat and Mouse -- Can Fake Text Generation Outpace Detector Systems?—0
Mitigating Hallucination of Large Vision-Language Models via Dynamic Logits CalibrationCode0
Uncovering Hidden Violent Tendencies in LLMs: A Demographic Analysis via Behavioral Vignettes—0
COIN: Uncertainty-Guarding Selective Question Answering for Foundation Models with Provable Risk Guarantees—0
Mastering Multiple-Expert Routing: Realizable H-Consistency and Strong Guarantees for Learning to Defer—0
Spotting Out-of-Character Behavior: Atomic-Level Evaluation of Persona Fidelity in Open-Ended GenerationCode0
Plan for Speed -- Dilated Scheduling for Masked Diffusion Language Models—0
LongWriter-Zero: Mastering Ultra-Long Text Generation via Reinforcement LearningCode5
OmniGen2: Exploration to Advanced Multimodal GenerationCode7
From General to Targeted Rewards: Surpassing GPT-4 in Open-Ended Long-Context Generation—0
GenerationPrograms: Fine-grained Attribution with Executable ProgramsCode0
S^4C: Speculative Sampling with Syntactic and Semantic Coherence for Efficient Inference of Large Language Models—0
Human-Centered Editable Speech-to-Sign-Language Generation via Streaming Conformer-Transformer and Resampling Hook—0
IGD: Token Decisiveness Modeling via Information Gain in LLMs for Personalized RecommendationCode0
Navigating the Black Box: Leveraging LLMs for Effective Text-Level Graph Injection Attacks—0
The Safety Reminder: A Soft Prompt to Reactivate Delayed Safety Awareness in Vision-Language Models—0
TagRouter: Learning Route to LLMs through Tags for Open-Domain Text Generation TasksCode1
Specification and Evaluation of Multi-Agent LLM Systems -- Prototype and Cybersecurity ApplicationsCode0
Reasoning RAG via System 1 or System 2: A Survey on Reasoning Agentic Retrieval-Augmented Generation for Industry ChallengesCode0
The Diffusion DualityCode3
OmniDRCA: Parallel Speech-Text Foundation Model via Dual-Resolution Speech Representations and Contrastive AlignmentCode0
Revisit What You See: Disclose Language Prior in Vision Tokens for Efficient Guided Decoding of LVLMsCode1
Adversarial Text Generation with Dynamic Contextual Perturbation—0
Evaluation empirique de la sécurisation et de l'alignement de ChatGPT et Gemini: analyse comparative des vulnérabilités par expérimentations de jailbreaks—0
Towards Secure and Private Language Models for Nuclear Power Plants—0
Diffuse Everything: Multimodal Diffusion Models on Arbitrary State SpacesCode1
ARGUS: Hallucination and Omission Evaluation in Video-LLMs—0
Eliciting Fine-Tuned Transformer Capabilities via Inference-Time TechniquesCode0
From Tool Calling to Symbolic Thinking: LLMs in a Persistent Lisp Metaprogramming Loop—0
HeavyWater and SimplexWater: Watermarking Low-Entropy Text DistributionsCode0
Towards Efficient Multi-LLM Inference: Characterization and Analysis of LLM Routing and Hierarchical Techniques—0
A Survey of Automatic Evaluation Methods on Text, Visual and Speech Generations—0
Controlling Summarization Length Through EOS Token Weighting—0
Do Large Language Models Judge Error Severity Like Humans?—0
Time to Talk: LLM Agents for Asynchronous Group Communication in Mafia GamesCode1
SuperWriter: Reflection-Driven Long-Form Generation with Large Language ModelsCode1
Show:102550
← PrevPage 1 of 107Next →
All datasetsDARTCOCO CaptionsEMNLP2017 WMTReDialCommonGenROCStoriesChinese PoemsCzech restaurant informationOpenWebTextSciQYahoo QuestionsADGEN

Benchmark Results

#ModelMetricClaimedVerifiedStatus
1T5B BaselineBLEU48.74—Unverified
2FactT5BBLEU48.37—Unverified
3JointGT BaselineBLEU47.51—Unverified
4FactJointGTBLEU47.39—Unverified
5Control Prefixes (T5-large)METEOR0.41—Unverified
6T5METEOR0.12—Unverified
7BARTMETEOR0.11—Unverified
#ModelMetricClaimedVerifiedStatus
1LeakGANBLEU-20.95—Unverified
2partGANBLEU-20.91—Unverified
3RankGANBLEU-20.85—Unverified
4RelGAN (100)BLEU-20.85—Unverified
5SeqGANBLEU-20.83—Unverified
#ModelMetricClaimedVerifiedStatus
1LeakGANBLEU-20.96—Unverified
2PPOGANBLEU-20.91—Unverified
3RelGANBLEU-20.88—Unverified
4SeqGANBLEU-20.86—Unverified
5RankGANBLEU-20.78—Unverified
#ModelMetricClaimedVerifiedStatus
1UniCRSDistinct-30.65—Unverified
2CRFRDistinct-30.52—Unverified
3KGSFDistinct-30.43—Unverified
4C2CRSDistinct-30.33—Unverified
5KBRDDistinct-30.3—Unverified
#ModelMetricClaimedVerifiedStatus
1UniLMCIDEr14.92—Unverified
2BART (TextBox 2.0)CIDEr12.98—Unverified
3BARTMETEOR0.3—Unverified
4T5METEOR0.29—Unverified
#ModelMetricClaimedVerifiedStatus
1Beam search + A*esque (sample)BLEU-134.4—Unverified
2Beam search + A*esque (beam)BLEU-134.4—Unverified
3Beam search + A*esque (greedy)BLEU-134.3—Unverified
4Beam searchBLEU-133.7—Unverified
#ModelMetricClaimedVerifiedStatus
1RankGANBLEU-20.81—Unverified
2SeqGANBLEU-20.74—Unverified
3LeakGANBLEU-20.46—Unverified
#ModelMetricClaimedVerifiedStatus
1TGen++METEOR0.17—Unverified
2TGenMETEOR0.15—Unverified
3TGen+METEOR0.15—Unverified
#ModelMetricClaimedVerifiedStatus
1GPT2-124Meval_loss3.12—Unverified
2GPT2-81M-LOOPeval_loss3.11—Unverified
3GPT2-Hermiteeval_loss2.91—Unverified
#ModelMetricClaimedVerifiedStatus
1LLaMA-65B+CFG (zero-shot)Accuracy96.6—Unverified
2LLaMA-30B+CFG (zero-shot)Accuracy96.4—Unverified
3LLaMA-13B+CFG (zero-shot)Accuracy95.1—Unverified
#ModelMetricClaimedVerifiedStatus
1CNN-VAENLL332.1—Unverified
2SA-VAENLL327.5—Unverified
3Aggressive VAENLL326.7—Unverified
#ModelMetricClaimedVerifiedStatus
1BART (TextBox 2.0)BLEU-410.2—Unverified
#ModelMetricClaimedVerifiedStatus
1STWGAN-GPBLEU-30.62—Unverified
#ModelMetricClaimedVerifiedStatus
1PALMROUGE-L41.41—Unverified
#ModelMetricClaimedVerifiedStatus
1BART (TextBox 2.0)ROUGE-L64.34—Unverified
#ModelMetricClaimedVerifiedStatus
1AEM+AttentionBLEU-114.17—Unverified
#ModelMetricClaimedVerifiedStatus
1GPT-4ASR65.1—Unverified
#ModelMetricClaimedVerifiedStatus
1BART (TextBox 2.0)ROUGE-L42.96—Unverified
#ModelMetricClaimedVerifiedStatus
1Graph2SeqBLEU22—Unverified
#ModelMetricClaimedVerifiedStatus
1WGANGP + DGflowJS-40.19—Unverified