Academic Publications: ACL Anthology (page 2)
Academic Publications · ACL Anthology: 52 entries in total; this page lists entries 16–30.
Beyond N-Grams: Rethinking Evaluation Metrics and Strategies for Multilingual Abstractive Summarization
Evidence grade: for discovery only; must not by itself support a formal conclusion.
The paper concerns computational linguistics and language technology. Its ACL 2025 Outstanding Paper award is verified, but independent external evidence of a major contribution to AI is still missing, so it is currently listed only as an entry awaiting further evidence.
Evidence use: For checking the paper's identity, authors, official award and original content; currently only a discovery and evidence-gathering entry point, not evidence of a major AI contribution confirmed by this site.
Limitations: Being listed publicly records a gap in the review; it does not mean the importance review was passed, nor that this site endorses all of the paper's conclusions. The official award shows recognition under that award process; until independent evidence of major contribution is added, it cannot be claimed that the paper significantly changed AI.
Link last checked: 2026-09-24
Typology-Guided Adaptation in Multilingual Models
Evidence grade: for discovery only; must not by itself support a formal conclusion.
The paper concerns foundation models and model architecture. Its ACL 2025 Outstanding Paper award is verified, but independent external evidence of a major contribution to AI is still missing, so it is currently listed only as an entry awaiting further evidence.
Evidence use: For checking the paper's identity, authors, official award and original content; currently only a discovery and evidence-gathering entry point, not evidence of a major AI contribution confirmed by this site.
Limitations: Being listed publicly records a gap in the review; it does not mean the importance review was passed, nor that this site endorses all of the paper's conclusions. The official award shows recognition under that award process; until independent evidence of major contribution is added, it cannot be claimed that the paper significantly changed AI.
Related concepts: Large Language Model (LLM) · Transformer
Link last checked: 2026-09-24
IndicSynth: A Large-Scale Multilingual Synthetic Speech Dataset for Low-Resource Indian Languages
Evidence grade: for discovery only; must not by itself support a formal conclusion.
The paper concerns computational linguistics and language technology. Its ACL 2025 Outstanding Paper award is verified, but independent external evidence of a major contribution to AI is still missing, so it is currently listed only as an entry awaiting further evidence.
Evidence use: For checking the paper's identity, authors, official award and original content; currently only a discovery and evidence-gathering entry point, not evidence of a major AI contribution confirmed by this site.
Limitations: Being listed publicly records a gap in the review; it does not mean the importance review was passed, nor that this site endorses all of the paper's conclusions. The official award shows recognition under that award process; until independent evidence of major contribution is added, it cannot be claimed that the paper significantly changed AI.
Link last checked: 2026-09-24
Towards the Law of Capacity Gap in Distilling Language Models
Evidence grade: for discovery only; must not by itself support a formal conclusion.
The paper concerns foundation models and model architecture. Its ACL 2025 Outstanding Paper award is verified, but independent external evidence of a major contribution to AI is still missing, so it is currently listed only as an entry awaiting further evidence.
Evidence use: For checking the paper's identity, authors, official award and original content; currently only a discovery and evidence-gathering entry point, not evidence of a major AI contribution confirmed by this site.
Limitations: Being listed publicly records a gap in the review; it does not mean the importance review was passed, nor that this site endorses all of the paper's conclusions. The official award shows recognition under that award process; until independent evidence of major contribution is added, it cannot be claimed that the paper significantly changed AI.
Related concepts: Large Language Model (LLM) · Transformer
Link last checked: 2026-09-24
Native Sparse Attention: Hardware-Aligned and Natively Trainable Sparse Attention
Evidence grade: for discovery only; must not by itself support a formal conclusion.
The paper concerns foundation models and model architecture; retrieval, agents and memory; and training and inference efficiency and infrastructure. Its ACL 2025 Best Paper award is verified, but independent external evidence of a major contribution to AI is still missing, so it is currently listed only as an entry awaiting further evidence.
Evidence use: For checking the paper's identity, authors, official award and original content; currently only a discovery and evidence-gathering entry point, not evidence of a major AI contribution confirmed by this site.
Limitations: Being listed publicly records a gap in the review; it does not mean the importance review was passed, nor that this site endorses all of the paper's conclusions. The official award shows recognition under that award process; until independent evidence of major contribution is added, it cannot be claimed that the paper significantly changed AI.
Related concepts: Large Language Model (LLM) · Transformer · Retrieval-Augmented Generation (RAG) · AI Agent · LLM Inference Optimization
Link last checked: 2026-09-24
Language Models Resist Alignment: Evidence From Data Compression
Evidence grade: for discovery only; must not by itself support a formal conclusion.
The paper concerns foundation models and model architecture, and evaluation, reliability and safety. Its ACL 2025 Best Paper award is verified, but independent external evidence of a major contribution to AI is still missing, so it is currently listed only as an entry awaiting further evidence.
Evidence use: For checking the paper's identity, authors, official award and original content; currently only a discovery and evidence-gathering entry point, not evidence of a major AI contribution confirmed by this site.
Limitations: Being listed publicly records a gap in the review; it does not mean the importance review was passed, nor that this site endorses all of the paper's conclusions. The official award shows recognition under that award process; until independent evidence of major contribution is added, it cannot be claimed that the paper significantly changed AI.
Related concepts: Large Language Model (LLM) · Transformer · Model Evaluation and Benchmarks
Link last checked: 2026-09-24
Capability Salience Vector: Fine-grained Alignment of Loss and Capabilities for Downstream Task Scaling Law
Evidence grade: for discovery only; must not by itself support a formal conclusion.
The paper concerns foundation models and model architecture, and evaluation, reliability and safety. Its ACL 2025 Outstanding Paper award is verified, but independent external evidence of a major contribution to AI is still missing, so it is currently listed only as an entry awaiting further evidence.
Evidence use: For checking the paper's identity, authors, official award and original content; currently only a discovery and evidence-gathering entry point, not evidence of a major AI contribution confirmed by this site.
Limitations: Being listed publicly records a gap in the review; it does not mean the importance review was passed, nor that this site endorses all of the paper's conclusions. The official award shows recognition under that award process; until independent evidence of major contribution is added, it cannot be claimed that the paper significantly changed AI.
Related concepts: Large Language Model (LLM) · Transformer · Model Evaluation and Benchmarks
Link last checked: 2026-09-24
LaTIM: Measuring Latent Token-to-Token Interactions in Mamba Models
Evidence grade: for discovery only; must not by itself support a formal conclusion.
The paper concerns foundation models and model architecture; retrieval, agents and memory; and evaluation, reliability and safety. Its ACL 2025 Outstanding Paper award is verified, but independent external evidence of a major contribution to AI is still missing, so it is currently listed only as an entry awaiting further evidence.
Evidence use: For checking the paper's identity, authors, official award and original content; currently only a discovery and evidence-gathering entry point, not evidence of a major AI contribution confirmed by this site.
Limitations: Being listed publicly records a gap in the review; it does not mean the importance review was passed, nor that this site endorses all of the paper's conclusions. The official award shows recognition under that award process; until independent evidence of major contribution is added, it cannot be claimed that the paper significantly changed AI.
Related concepts: Large Language Model (LLM) · Transformer · Retrieval-Augmented Generation (RAG) · AI Agent · Model Evaluation and Benchmarks
Link last checked: 2026-09-24
LLMs know their vulnerabilities: Uncover Safety Gaps through Natural Distribution Shifts
Evidence grade: for discovery only; must not by itself support a formal conclusion.
The paper concerns foundation models and model architecture, and evaluation, reliability and safety. Its ACL 2025 Outstanding Paper award is verified, but independent external evidence of a major contribution to AI is still missing, so it is currently listed only as an entry awaiting further evidence.
Evidence use: For checking the paper's identity, authors, official award and original content; currently only a discovery and evidence-gathering entry point, not evidence of a major AI contribution confirmed by this site.
Limitations: Being listed publicly records a gap in the review; it does not mean the importance review was passed, nor that this site endorses all of the paper's conclusions. The official award shows recognition under that award process; until independent evidence of major contribution is added, it cannot be claimed that the paper significantly changed AI.
Related concepts: Large Language Model (LLM) · Transformer · Model Evaluation and Benchmarks
Link last checked: 2026-09-24
All That Glitters is Not Novel: Plagiarism in AI Generated Research
Evidence grade: for discovery only; must not by itself support a formal conclusion.
The paper concerns foundation models and model architecture. Its ACL 2025 Outstanding Paper award is verified, but independent external evidence of a major contribution to AI is still missing, so it is currently listed only as an entry awaiting further evidence.
Evidence use: For checking the paper's identity, authors, official award and original content; currently only a discovery and evidence-gathering entry point, not evidence of a major AI contribution confirmed by this site.
Limitations: Being listed publicly records a gap in the review; it does not mean the importance review was passed, nor that this site endorses all of the paper's conclusions. The official award shows recognition under that award process; until independent evidence of major contribution is added, it cannot be claimed that the paper significantly changed AI.
Related concepts: Large Language Model (LLM) · Transformer
Link last checked: 2026-09-24
I0T: Embedding Standardization Method Towards Zero Modality Gap
Evidence grade: for discovery only; must not by itself support a formal conclusion.
The paper concerns computational linguistics and language technology. Its ACL 2025 Outstanding Paper award is verified, but independent external evidence of a major contribution to AI is still missing, so it is currently listed only as an entry awaiting further evidence.
Evidence use: For checking the paper's identity, authors, official award and original content; currently only a discovery and evidence-gathering entry point, not evidence of a major AI contribution confirmed by this site.
Limitations: Being listed publicly records a gap in the review; it does not mean the importance review was passed, nor that this site endorses all of the paper's conclusions. The official award shows recognition under that award process; until independent evidence of major contribution is added, it cannot be claimed that the paper significantly changed AI.
Link last checked: 2026-09-24
Rethinking the Role of Prompting Strategies in LLM Test-Time Scaling: A Perspective of Probability Theory
Evidence grade: for discovery only; must not by itself support a formal conclusion.
The paper concerns foundation models and model architecture, and training, alignment and reasoning. Its ACL 2025 Outstanding Paper award is verified, but independent external evidence of a major contribution to AI is still missing, so it is currently listed only as an entry awaiting further evidence.
Evidence use: For checking the paper's identity, authors, official award and original content; currently only a discovery and evidence-gathering entry point, not evidence of a major AI contribution confirmed by this site.
Limitations: Being listed publicly records a gap in the review; it does not mean the importance review was passed, nor that this site endorses all of the paper's conclusions. The official award shows recognition under that award process; until independent evidence of major contribution is added, it cannot be claimed that the paper significantly changed AI.
Related concepts: Large Language Model (LLM) · Transformer · Fine-tuning · Reasoning Models
Link last checked: 2026-09-24
PARME: Parallel Corpora for Low-Resourced Middle Eastern Languages
Evidence grade: for discovery only; must not by itself support a formal conclusion.
The paper concerns computational linguistics and language technology. Its ACL 2025 Outstanding Paper award is verified, but independent external evidence of a major contribution to AI is still missing, so it is currently listed only as an entry awaiting further evidence.
Evidence use: For checking the paper's identity, authors, official award and original content; currently only a discovery and evidence-gathering entry point, not evidence of a major AI contribution confirmed by this site.
Limitations: Being listed publicly records a gap in the review; it does not mean the importance review was passed, nor that this site endorses all of the paper's conclusions. The official award shows recognition under that award process; until independent evidence of major contribution is added, it cannot be claimed that the paper significantly changed AI.
Link last checked: 2026-09-24
A Theory of Response Sampling in LLMs: Part Descriptive and Part Prescriptive
Evidence grade: for discovery only; must not by itself support a formal conclusion.
The paper concerns foundation models and model architecture. Its ACL 2025 Best Paper award is verified, but independent external evidence of a major contribution to AI is still missing, so it is currently listed only as an entry awaiting further evidence.
Evidence use: For checking the paper's identity, authors, official award and original content; currently only a discovery and evidence-gathering entry point, not evidence of a major AI contribution confirmed by this site.
Limitations: Being listed publicly records a gap in the review; it does not mean the importance review was passed, nor that this site endorses all of the paper's conclusions. The official award shows recognition under that award process; until independent evidence of major contribution is added, it cannot be claimed that the paper significantly changed AI.
Related concepts: Large Language Model (LLM) · Transformer
Link last checked: 2026-09-24
Revisiting Compositional Generalization Capability of Large Language Models Considering Instruction Following Ability
Evidence grade: for discovery only; must not by itself support a formal conclusion.
The paper concerns foundation models and model architecture. Its ACL 2025 Outstanding Paper award is verified, but independent external evidence of a major contribution to AI is still missing, so it is currently listed only as an entry awaiting further evidence.
Evidence use: For checking the paper's identity, authors, official award and original content; currently only a discovery and evidence-gathering entry point, not evidence of a major AI contribution confirmed by this site.
Limitations: Being listed publicly records a gap in the review; it does not mean the importance review was passed, nor that this site endorses all of the paper's conclusions. The official award shows recognition under that award process; until independent evidence of major contribution is added, it cannot be claimed that the paper significantly changed AI.
Related concepts: Large Language Model (LLM) · Transformer
Link last checked: 2026-09-24