Challenge: EvAlign is a visual analytics framework for quantitative and qualitative evaluation of automatic translation alignment models.
Approach: They propose to use EvAlign to analyze automatic translation alignment models and compare their performance with other baseline and state-of-the-art models.
Outcome: The framework hosts nine gold standard datasets and the predictions of multiple alignment models.

Similar Papers

SpeechAlign: A Framework for Speech Translation Alignment Evaluation (2024.lrec-main)

Copied to clipboard

Challenge: Speech-to-Speech and Speech- to-Text translation are currently dynamic areas of research.
Approach: They propose a framework to evaluate source-target alignment in speech models . they introduce a speech gold alignment dataset and introduce two new metrics .
Outcome: The proposed framework evaluates source-target alignment quality within speech models.
AutoAlign: Get Your LLM Aligned with Minimal Annotations (2025.acl-demo)

Copied to clipboard

Challenge: Automated Alignment (ALM) is a set of algorithms designed to align Large Language Models (LLMs) with human intentions and values while minimizing manual intervention.
Approach: They propose an open-source toolkit that integrates mainstream automated algorithms through a consistent interface and an accessible workflow supporting one-click execution for prompt synthesis and automatic alignment signal construction.
Outcome: The proposed framework enables easy reproduction of existing results through extensive benchmarks and facilitates the development of novel approaches via modular components.
SilverAlign: MT-Based Silver Data Algorithm for Evaluating Word Alignment (2024.lrec-main)

Copied to clipboard

Challenge: Word alignments are crucial for a variety of NLP tasks.
Approach: They propose a method to automatically create silver data for evaluation of word aligners by exploiting machine translation and minimal pairs.
Outcome: The proposed method correlates with gold benchmarks for 9 language pairs, making it a valid resource for evaluation of different languages and domains when gold data is not available.
AlignMMBench: Evaluating Chinese Multimodal Alignment in Large Vision-Language Models (2025.acl-long)

Copied to clipboard

Challenge: Existing benchmarks focus on basic abilities using nonverbal methods, such as yes-no and multiple-choice questions.
Approach: They propose a benchmark that provides more nuanced evaluations of alignment capabilities for large Vision-Language Models (VLMs) they use a rule-calibrated evaluator that exceeds GPT-4's evaluation ability and a “alignment score” to assess the robustness and stability of models across diverse prompts.
Outcome: The proposed benchmark covers 13 tasks across three categories and includes both single-turn and multi-turn dialogue scenarios.
Large Vision-Language Model Alignment and Misalignment: A Survey Through the Lens of Explainability (2025.findings-emnlp)

Copied to clipboard

Challenge: Large Vision-Language Models have demonstrated remarkable capabilities in processing both visual and textual information.
Approach: They examine the challenge of alignment and misalignment in LVLMs through an explainability lens.
Outcome: The findings highlight the need for standardized evaluation protocols and in-depth explainability studies.
MetaAlign: Align Large Language Models with Diverse Preferences during Inference Time (2025.findings-naacl)

Copied to clipboard

Challenge: Existing methods to align large language models with human preferences often result in a static alignment that cannot account for the diversity of human preferences in practical applications.
Approach: They propose a method to help large language models dynamically align with various explicit or implicit preferences specified at inference time.
Outcome: The proposed method can help LLMs dynamically align with various explicit or implicit preferences specified at the inference stage, validating the feasibility of MetaAlign.
Understanding Cross-Lingual Alignment—A Survey (2024.findings-acl)

Copied to clipboard

Challenge: Cross-lingual alignment is the meaningful similarity of representations across languages in multilingual language models.
Approach: They propose a taxonomy of methods to improve cross-lingual alignment . they argue that an effective trade-off between language-neutral and language-specific information is key .
Outcome: The proposed methods can be applied to encoder models and encoder-decoder-only models . they show that language-neutral and language-specific information is key .
Exploring and Evaluating Attributes, Values, and Structures for Entity Alignment (2020.emnlp-main)

Copied to clipboard

Challenge: Entity alignment (EA) aims at building a Knowledge Graph (KG) of rich content by linking the equivalent entities from various KGs.
Approach: They propose to use an attributed value encoder to partition a Knowledge Graph into subgraphs to model the various types of attribute triples efficiently.
Outcome: The proposed method achieves significant improvements over 12 baselines in cross-lingual and monolingual datasets.
BinaryAlign: Word Alignment as Binary Sequence Labeling (2024.acl-long)

Copied to clipboard

Challenge: State-of-the-art word alignment training methods require a different class depending on the availability of gold data for a particular language pair.
Approach: They propose a novel word alignment technique based on binary sequence labeling that outperforms existing approaches in both scenarios.
Outcome: The proposed method outperforms existing models on non-English language pairs and performs stratified error analysis over alignment error type.
AlignBench: Benchmarking Chinese Alignment of Large Language Models (2024.acl-long)

Copied to clipboard

Challenge: Effective evaluation of alignment for emerging Chinese LLMs is still significantly lacking, calling for real-scenario grounded, open-ended, challenging and automatic evaluations tailored for alignment.
Approach: They propose a multi-dimensional benchmark for evaluating LLMs’ alignment in Chinese with 8 main categories, 683 real-scenario rooted queries and corresponding human verified references.
Outcome: The benchmark uses a human-in-the-loop data curation pipeline, 683 real-scenario rooted queries and human verified references.

What is GenGO?

GenGO is an NLP powered publication search system. It currenctly indexes 30k+ papers from ACL Anthology, and implements multi-aspect summarization, semantic search, and more!

Information

About
Limitations