REVIEW 15 cited by
Contrastive Preference Optimization: Pushing the Boundaries of LLM Performance in Machine Translation
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
read the original abstract
Moderate-sized large language models (LLMs) -- those with 7B or 13B parameters -- exhibit promising machine translation (MT) performance. However, even the top-performing 13B LLM-based translation models, like ALMA, does not match the performance of state-of-the-art conventional encoder-decoder translation models or larger-scale LLMs such as GPT-4. In this study, we bridge this performance gap. We first assess the shortcomings of supervised fine-tuning for LLMs in the MT task, emphasizing the quality issues present in the reference data, despite being human-generated. Then, in contrast to SFT which mimics reference translations, we introduce Contrastive Preference Optimization (CPO), a novel approach that trains models to avoid generating adequate but not perfect translations. Applying CPO to ALMA models with only 22K parallel sentences and 12M parameters yields significant improvements. The resulting model, called ALMA-R, can match or exceed the performance of the WMT competition winners and GPT-4 on WMT'21, WMT'22 and WMT'23 test datasets.
Forward citations
Cited by 15 Pith papers
-
LPOI: Listwise Preference Optimization for Vision Language Models
LPOI reduces VLM hallucination by training the model to prefer the original image over progressively masked versions of the same image, using a listwise ranking loss built from pairwise preference data.
-
Every Sample Counts: Supervised Fine-Tuning of Language Models with Pointwise Constraints
Pointwise constrained fine-tuning via sample-wise augmented Lagrangians and learned relaxations reduces tail constraint violations across safety, tool-calling, and re-ranking while preserving average task performance.
-
Epistemic-aware Vision-Language Foundation Model for Fetal Ultrasound Interpretation
A multi-view vision-language model trained on 20,000 fetal ultrasound reports generates clinical text and diagnoses, reportedly outperforming general and medical baselines.
-
Quality-Aware Decoding: Unifying Quality Estimation and Decoding
A uni-directional token-level QE model that scores partial translations is merged into beam search, improving NMT quality over N-best re-ranking on WMT23 English-German and Chinese-English.
-
AIGB-R1: Self-Evolving Generative Auto-Bidding via Hierarchical Planner-Executor Optimization
AIGB-R1 couples an LLM strategy planner with a prompt-conditioned Decision Transformer executor and trains both end-to-end with a decoupled-GRPO self-evolving loop, reporting the highest Score on the AuctionNet benchmark.
-
$M^2PO$: Multi-Perspective Multi-Pair Preference Optimization for Machine Translation
M2PO combines a QE-plus-alignment reward with a dynamic curriculum and multi-pair DPO loss, and reports WMT21-22 gains for a 7B translation model, but the abstract's WMT23/24 9B parity claims are unsupported.
-
Align-then-Slide: A complete evaluation framework for Ultra-Long Document-Level Machine Translation
An evaluation framework for document-level MT that aligns source and target sentences via dynamic programming and averages 1-4 chunk sliding-window scores, reporting 0.929 Pearson correlation with human MQM and enabli...
-
RIVAL: Reinforcement Learning with Iterative and Adversarial Optimization for Machine Translation
RIVAL iteratively re-trains a reward model adversarially against the current translator and adds a BLEU-predicting head, improving in-domain WMT and subtitle translation over SFT baselines.
-
AUTOLAW: Enhancing Legal Compliance in Large Language Models via Case Law Generation and Jury-Inspired Deliberation
AutoLaw's verifier-ranked legal-role jury with a similar-case demonstration beats majority voting for violation detection on three law and policy benchmarks.
-
Beyond the Sentence: A Survey on Context-Aware Machine Translation with Large Language Models
A survey of context-aware machine translation with large language models, categorizing prompting, fine-tuning, and agent-based approaches.
-
Dynamic Chunking and Selection for Reading Comprehension of Ultra-Long Context in Large Language Models
Dynamic chunking plus question-aware chunk selection improves long-context QA, but the headline numbers are partly inflated by choosing hyperparameters on the test benchmarks.
-
Modeling and Optimizing User Preferences in AI Copilots: A Comprehensive Survey and Taxonomy
AI copilot preference optimization is organized into a pre-, mid-, and post-interaction taxonomy, with a unified definition of AI copilots.
-
Rethinking the Outlier Distribution in Large Language Models: An In-depth Study
The paper claims that most massive activations in LLMs are residual-propagated 'fake' outliers that can be removed without accuracy loss.
-
Redefining Simplicity: Benchmarking Large Language Models from Lexical to Document Simplification
In a four-task benchmark, GPT-4o, Llama3.1-70B, and Gemma2-2B outperform traditional text simplification systems on most automatic metrics, and GPT-4o is preferred over human-written references in a small human study.
-
DPO-Shift: Shifting the Distribution of Direct Preference Optimization
DPO-Shift multiplies the rejected-response term in the DPO loss by a factor f(λ) below 1, shifting chosen-response probabilities upward at the expense of some reward margin.
Discussion (0). Continue with ORCID to comment.