Papers with FOCUS
Fast or Slow? Integrating Fast Intuition and Deliberate Thinking for Enhancing Visual Question Answering (2025.acl-short)
Copied to clipboard
| Challenge: | Current approaches generate visual markers for all questions, generating excessive visual markers. |
| Approach: | They propose a plug-and-play approach that adapts to the complexity of questions . they propose combining fast intuitive judgments with deliberate analytical reasoning . |
| Outcome: | The proposed approach improves performance on four benchmarks on ScienceQA, TextQA, VizWiz, and MME. |
Evaluating Tokenizer Adaptation Methods for Large Language Models on Low-Resource Programming Languages (2025.acl-srw)
Copied to clipboard
| Challenge: | Large language models (LLMs) trained on high-resource programming languages perform sub-optimally for low-resourced programming languages (LRPLs). |
| Approach: | They evaluate the impact of tokenizer adaptation methods on improving code generation for LRPLs. |
| Outcome: | The proposed methods outperform the original models and fine-tuned models in LRPLs, but performance declines in non-target languages like Python after tokenizer adaptation. |
EnerGIZAr: Leveraging GIZA++ for Effective Tokenizer Initialization (2025.findings-acl)
Copied to clipboard
| Challenge: | Continual pre-training has long been considered the default strategy for adapting models to non-English languages, but struggles with initializing new embeddings, especially for non-Latin scripts. |
| Approach: | They propose a method that leverages statistical word alignment techniques to improve continual pre-training by leveraging word alignment matrix between source and target tokens. |
| Outcome: | The proposed method outperforms existing methods on key NLP tasks including POS tagging, Sentiment Analysis, NLI, and NER in Hindi, Basque, Arabic and Korean. |
SLANG: New Concept Comprehension of Large Language Models (2024.emnlp-main)
Copied to clipboard
| Challenge: | Dynamic nature of language limits the adaptability of Large Language Models (LLMs) Traditionally, LLMs are trained on static data, which limits their adaptability . |
| Approach: | They propose a benchmark to integrate novel data and assess LLMs’ ability to comprehend emerging concepts, alongside a causal inference-based approach to enhance LLM comprehension of new phrases and their colloquial context. |
| Outcome: | The proposed model outperforms baseline models in terms of precision and relevance in the comprehension of Internet slang and memes. |
FOCUS: Effective Embedding Initialization for Monolingual Specialization of Multilingual Models (2023.emnlp-main)
Copied to clipboard
| Challenge: | Multilingual models have been released, but many of the world's languages are not covered. |
| Approach: | They propose a method that initializes the embedding matrix for a new tokenizer based on information in the source model's embeddable matrix. |
| Outcome: | The proposed method outperforms random initialization and previous work on language modeling and on a range of downstream tasks (NLI, QA, and NER). |
FOCUS: A Fine-Grained Customer-Oriented Sentiment Dialogue Summarization Dataset for Chinese Customer Service (2026.findings-acl)
Copied to clipboard
| Challenge: | Existing studies largely overlook fine-grained sentiment dynamics expressed by customers . current methods often exhibit misalignment between aspects and sentiments . |
| Approach: | They propose a three-stage approach to building an aspect-aware sentiment dataset . they use a fine-grained customer-oriented Chinese dialogUe summarization dataset based on this scheme . |
| Outcome: | The proposed model improves faithfulness and interpretability of the proposed dataset. |
FOCUS: Evaluating Pre-trained Vision-Language Models on Underspecification Reasoning (2025.acl-long)
Copied to clipboard
| Challenge: | a new dataset evaluates whether vision-language models have underspecification reasoning abilities . underspecifications are often left incomplete or vague, and are often ignored for mutual understanding . |
| Approach: | They propose a probing dataset to evaluate whether VLMs have underspecification reasoning . they find that pre-trained vision-language models lack this ability . |
| Outcome: | The proposed probing dataset shows that pre-trained vision-language models lack underspecification reasoning abilities. |
Measuring Social Bias in Vision-Language Models with Face-Only Counterfactuals from Real Photos (2026.acl-long)
Copied to clipboard
| Challenge: | Vision-Language Models (VLMs) are increasingly deployed in socially consequential settings . attribution under visual confounding is a central challenge in measuring social bias . |
| Approach: | They propose a face-only counterfactual evaluation paradigm that isolates demographic effects while preserving real-image realism. |
| Outcome: | The proposed paradigm isolates demographic effects while preserving real-image realism. |