Sequential Fine-Tuning Approach Improves Automated Essay Scoring by Modeling Discourse Structure
Researchers found that fine-tuning LLaMA-3.1-8B sequentially on essay components (lead, position, claim, evidence, conclusion) in order produces better automated essay scoring than training on these elements independently or randomly. The sequential approach achieved F1 scores of 65% for evidence and 87% for conclusion, outperforming a much larger LLaMA-70B baseline on some tasks. This suggests that smaller, task-optimized language models can match larger models' performance when trained with curricula aligned to the structure of the task.
A new study on automated essay scoring (AES) demonstrates that the order of fine-tuning matters significantly when training language models to evaluate essays. Researchers compared three training approaches using LLaMA-3.1-8B with parameter-efficient LoRA and 4-bit quantization on the PERSUADE 2.0 corpus: sequential fine-tuning on discourse elements in logical order, independent task-specific models, and randomized multi-task training. Sequential fine-tuning—progressively training on lead, position, claim, evidence, and conclusion—produced the strongest overall results, with F1 scores of 65% for evidence and 87% for conclusion. Notably, this smaller model outperformed a general-purpose LLaMA-70B baseline on conclusion scoring despite having far fewer parameters. The findings indicate that curriculum design aligned with discourse structure can materially improve AES performance and that cost-effective, smaller models can be competitive with substantially larger language models when properly optimized.
What's missing
The study does not discuss potential limitations of the PERSUADE 2.0 corpus (e.g., essay types, grade levels, or demographic representation), nor does it address how the approach generalizes to essays outside this dataset or to non-English languages. The paper also does not compare against other state-of-the-art AES systems beyond the LLaMA-70B baseline.
What different sources said
- arXiv cs.LGCenter
The Order Matters: Sequential Fine-Tuning of LLaMA for Coherent Automated Essay Scoring
Related
Gut Bacteria Enzyme Found to Break Down Heat-Processed Food Compounds, Producing Novel Biogenic Amines
Researchers have discovered that an enzyme in common gut bacteria can degrade N-epsilon-carboxymethyllysine (CML), a compound formed during thermal food processing, producing previously unknown biogenic amines. The enzyme, ornithine decarboxylase SpeC from enterobacteria, acts on CML and related modified lysine derivatives through a low-level 'underground' catalytic activity. This finding suggests a previously unrecognized communication axis between thermally processed dietary compounds and gut microbial physiology, with potential implications for host health.
Full-Length Gene Sequencing Reveals Two Distinct Bacterial Communities in Black-Legged Ticks Expanding Into Canada
Researchers used Oxford Nanopore full-length 16S rRNA gene sequencing to characterize the microbiome of Ixodes scapularis black-legged ticks collected in Nova Scotia, Canada, distinguishing between tick-adapted bacteria and environmentally acquired bacteria. The study comes as I. scapularis — the primary vector of Lyme disease — is rapidly expanding northward into Canada due to climate change. The findings suggest that environmentally derived bacteria in tick microbiomes are not mere contamination, which has implications for how tick microbiome data is collected and interpreted across surveillance studies.
Study Identifies Metabolic Link Between Cell Envelope Stress and Biofilm Formation in Bacteria
Researchers have discovered that the metabolite acetyl-CoA directly inhibits enzymes that degrade the bacterial signaling molecule c-di-GMP, connecting cell envelope biosynthesis stress to biofilm formation in Pseudomonas aeruginosa. The study found that sub-inhibitory concentrations of antibiotics targeting early peptidoglycan biosynthesis — but not other antibiotic classes — elevate c-di-GMP levels by reducing phosphodiesterase activity, with acetyl-CoA competing for the enzyme active site. Because the relevant enzyme domain is broadly conserved across bacterial species, this checkpoint mechanism may be widespread and could have implications for understanding antibiotic-induced biofilm responses.