REVIEW 10 cited by
Can LLMs Effectively Leverage Graph Structural Information through Prompts, and Why?
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
Large language models (LLMs) are gaining increasing attention for their capability to process graphs with rich text attributes, especially in a zero-shot fashion. Recent studies demonstrate that LLMs obtain decent text classification performance on common text-rich graph benchmarks, and the performance can be improved by appending encoded structural information as natural languages into prompts. We aim to understand why the incorporation of structural information inherent in graph data can improve the prediction performance of LLMs. First, we rule out the concern of data leakage by curating a novel leakage-free dataset and conducting a comparative analysis alongside a previously widely-used dataset. Second, as past work usually encodes the ego-graph by describing the graph structure in natural language, we ask the question: do LLMs understand the graph structure in accordance with the intent of the prompt designers? Third, we investigate why LLMs can improve their performance after incorporating structural information. Our exploration of these questions reveals that (i) there is no substantial evidence that the performance of LLMs is significantly attributed to data leakage; (ii) instead of understanding prompts as graph structures as intended by the prompt designers, LLMs tend to process prompts more as contextual paragraphs and (iii) the most efficient elements of the local neighborhood included in the prompt are phrases that are pertinent to the node label, rather than the graph structure.
Forward citations
Cited by 10 Pith papers
-
Agentic Graph Token Reasoning
Agentic Graph Token (AGT) reasoning lets an LLM iteratively select graph views, encode them as tokens, and reason step-by-step, beating static graph-token and text-agent baselines on seven graph domains.
-
Beyond the Golden Teacher: Enhancing Graph Learning through LLM-GNN Co-teaching
Bidirectional LLM-GNN co-teaching with round-based pseudo-label preference optimization outperforms golden-teacher baselines on few-shot TAG benchmarks by 3-8% absolute gains.
-
Clustering as Reasoning: A $k$-Means Interpretation of Chain-of-Thought Graph Learning
KCoT reframes CoT graph learning as k-means clustering by establishing a formal correspondence between Transformer blocks and k-means assignment/update steps, with a Semantic Discriminating Prompt and structure alignm...
-
Revisiting Graph-Tokenizing Large Language Models: A Systematic Evaluation of Graph Token Understanding
GTokenLLMs do not fully understand graph tokens, exhibiting over-sensitivity or insensitivity to instruction changes and relying heavily on text for reasoning even when graph information is preserved.
-
Court of LLMs: Evidence-Augmented Generation via Multi-LLM Collaboration for Text-Attributed Graph Anomaly Detection
CoLL uses two specialized LLM 'prosecutors' and an LLM 'judge' to generate textual anomaly evidence, which a gated GNN then fuses with graph structure for state-of-the-art text-attributed graph anomaly detection.
-
AGE: Adaptive-masking for Graph Embedding in Graph Retrieval-Augmented Generation
AGE applies adaptive masking via a learnable sampler in Transformer-based SSL to align graph and text embeddings, yielding higher accuracy on four GraphQA benchmarks for non-parametric GraphRAG.
-
GLIP: Graph and LLM Joint Pretraining for Graph-Level Tasks
GLIP is a joint GNN-LLM pretraining framework that uses augmentation, multi-token selection, a diffusion projector, and combined contrastive plus semantic losses to boost graph classification and reasoning after fine-...
-
Knowledge Graph-Enhanced Zero-Shot Topic Classification: A Multi-Strategy Comparative Study
Keyword-enhanced classification performs best among zero-shot variants; knowledge graph augmentation improves small models but degrades large ones, while self-consistency adds cost without benefit.
-
STK-Adapter: Incorporating Evolving Graph and Event Chain for Temporal Knowledge Graph Extrapolation
STK-Adapter adds Spatial-Temporal MoE, Event-Aware MoE, and Cross-Modality Alignment MoE to integrate evolving TKG graphs and event chains into LLMs, reducing information loss and improving extrapolation performance o...
-
A Graph-Enhanced Defense Framework for Explainable Fake News Detection with LLM
G-Defense builds claim-centered graphs from sub-claims, applies RAG for evidence and competing explanations, then uses graph inference to detect fake news veracity and generate intuitive explanation graphs, claiming S...
Discussion (0). Sign in to comment.