Graph Memory Transformer (GMT)

๐Ÿ“… 2026-04-26
๐Ÿ“ˆ Citations: 0
โœจ Influential: 0
๐Ÿ“„ PDF
๐Ÿค– AI Summary
This work proposes the Graph Memory Transformer (GMT), which replaces the feed-forward networks (FFNs) in a standard autoregressive Transformer decoder with an explicit, learnable memory graph while preserving the causal self-attention mechanism. GMT implements interpretable state transitions through a routing-and-displacement scheme based on 128 memory centroids, a 128ร—128 directed transition matrix, gravity-source routing, token-conditioned target selection, and gated readout, yielding a pure decoder language model without FFNs. With 82.2M parameters, GMT trains stably and achieves validation loss and perplexity slightly behind those of a 103.0M-parameter GPT baseline (3.5995/36.58 vs. 3.2903/26.85), yet demonstrates comparable zero-shot performance, thereby validating the feasibility and interpretability of the graph-based memory mechanism.

Technology Category

Application Category

๐Ÿ“ Abstract
We investigate whether the Feed-Forward Network (FFN) sublayer in a decoder-only transformer can be replaced by an explicit learned memory graph while preserving the surrounding autoregressive architecture. The proposed Graph Memory Transformer (GMT) keeps causal self-attention intact, but replaces the usual per-token FFN transformation with a memory cell that routes token representations over a learned bank of centroids connected by a learned directed transition matrix. In the base GMT v7 instantiation studied here, each of 16 transformer blocks contains 128 centroids, a 128 * 128 edge matrix, gravitational source routing, token-conditioned target selection, and a gated displacement readout. The cell therefore returns movement from an estimated source memory state toward a target memory state, rather than a retrieved value. The resulting model is a fully decoder-only language model with 82.2M trainable parameters and no dense FFN sublayers, compared with a 103.0M-parameter dense GPT-style baseline used in the evaluation. The base v7 model trains stably and exposes centroid usage, transition structure, and source-to-target movement as directly inspectable quantities of the forward computation. It remains behind the larger dense baseline in validation loss and perplexity (3.5995/36.58 vs. 3.2903/26.85), while showing close zero-shot benchmark behavior under the evaluated setting. These results are not intended as a state-of-the-art claim; they support the viability and structural interpretability of replacing dense within-token transformation with graph-mediated memory navigation. Broader scaling, optimized kernels, and more extensive benchmark evaluation are left for subsequent work.
Problem

Research questions and friction points this paper is trying to address.

Graph Memory Transformer
Feed-Forward Network replacement
memory graph
decoder-only transformer
autoregressive architecture
Innovation

Methods, ideas, or system contributions that make the work stand out.

Graph Memory Transformer
memory graph
centroid routing
decoder-only architecture
structural interpretability
๐Ÿ”Ž Similar Papers
2024-07-13arXiv.orgCitations: 36