Papers › Investigating Efficiently Extending Transformers for Long Input Summarization

Investigating Efficiently Extending Transformers for Long Input Summarization

8 Aug 2022arXiv:2208.04347archive 2025-07-28

Jason Phang, Yao Zhao, Peter J. Liu

While large pretrained Transformer models have proven highly capable at tackling natural language tasks, handling long sequence inputs continues to be a significant challenge. One such task is long input summarization, where inputs are longer than the maximum input context of most pretrained models. Through an extensive set of experiments, we investigate what model architectural changes and pretraining paradigms can most efficiently adapt a pretrained Transformer for long input summarization. We find that a staggered, block-local Transformer with global encoder tokens strikes a good balance of performance and efficiency, and that an additional pretraining phase on long sequences meaningfully improves downstream summarization performance. Based on our findings, we introduce PEGASUS-X, an extension of the PEGASUS model with additional long input pretraining to handle inputs of up to 16K tokens. PEGASUS-X achieves strong performance on long input summarization tasks comparable with much larger models while adding few additional parameters and not requiring model parallelism to train.

PaperPDFCode

In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

Code

google-research/pegasus officialmentioned in papertfApache-2.0 report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

16kLong-range modelingText Summarization

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Long-range modeling SCROLLS PEGASUS-X GovRep 60.3 / 30.0 / 31.5 #11 of 13 Archive leaderboard report
Long-range modeling SCROLLS PEGASUS-X QMSum 33.2 / 9.6 / 21.6 #11 of 13 Archive leaderboard report
Long-range modeling SCROLLS PEGASUS-X SumScr 35.7 / 9.1 / 20.6 #11 of 13 Archive leaderboard report
Long-range modeling SCROLLS PEGASUS-X-Base GovRep 59.3 / 29.3 / 30.9 #12 of 13 Archive leaderboard report
Long-range modeling SCROLLS PEGASUS-X-Base QMSum 32.9 / 9.8 / 21.4 #12 of 13 Archive leaderboard report
Long-range modeling SCROLLS PEGASUS-X-Base SumScr 35.0 / 8.9 / 20.4 #12 of 13 Archive leaderboard report
Text Summarization Arxiv HEP-TH citation graph Pegasus-X ROUGE-1 50.0 #3 of 28 Archive leaderboard report
Text Summarization Arxiv HEP-TH citation graph Pegasus-X ROUGE-2 21.8 #3 of 28 Archive leaderboard report
Text Summarization Arxiv HEP-TH citation graph Pegasus-X ROUGE-L 44.6 #3 of 28 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Methods

Absolute Position EncodingsAdamAttentionBPEDense ConnectionsDropoutLabel SmoothingLayer NormalizationLinear LayerMulti-Head AttentionPEGASUSPosition-Wise Feed-Forward LayerResidual ConnectionSoftmaxTransformer

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections