diff --git a/NAMESPACE b/NAMESPACE
index 8d56a05..6e15caf 100644
--- a/NAMESPACE
+++ b/NAMESPACE
@@ -319,6 +319,7 @@ importFrom(methods,slot)
 importFrom(methods,slotNames)
 importFrom(methods,validObject)
 importFrom(rjson,fromJSON)
+importFrom(rlang,"%||%")
 importFrom(rlang,check_installed)
 importFrom(sf,"st_agr<-")
 importFrom(sf,"st_geometry<-")
diff --git a/R/aggregate.R b/R/aggregate.R
index bf9f8b5..a119178 100644
--- a/R/aggregate.R
+++ b/R/aggregate.R
@@ -374,6 +374,7 @@ aggregateTxTech <- function(data_dir, df = NULL, by = NULL,
 #'
 #' @inheritParams sf::st_make_grid
 #' @inheritParams sf::aggregate
+#' @inheritParams aggregateTx
 #' @param x An SFE object to be aggregated.
 #' @param by A \code{sf} data frame whose geometry column is used for
 #'   aggregation or \code{sfc} or for multiple samples a list of \code{sfc}
diff --git a/man/aggregate-SpatialFeatureExperiment-method.Rd b/man/aggregate-SpatialFeatureExperiment-method.Rd
index 774fc00..09032b8 100644
--- a/man/aggregate-SpatialFeatureExperiment-method.Rd
+++ b/man/aggregate-SpatialFeatureExperiment-method.Rd
@@ -59,6 +59,11 @@ output. Defaults to "bins".}
 
 \item{join}{logical spatial predicate function to use if \code{by} is a simple features object or geometry; see \link[sf]{st_join}}
 
+\item{sparse}{Logical, whether the gene count matrix from aggregating
+transcript spots should be sparse. When the bins are large, the matrix will
+not be very sparse so using sparse matrix will not save memory, but when
+the bins are small, sparsity is worth it.}
+
 \item{BPPARAM}{A \code{\link{BiocParallelParam}} object specifying parallel
 computing when aggregating data with functions other than sum and mean when
 aggregating cells. When aggregating transcript spots, this specifies
diff --git a/man/aggregateTx.Rd b/man/aggregateTx.Rd
index 3cb279b..c72c504 100644
--- a/man/aggregateTx.Rd
+++ b/man/aggregateTx.Rd
@@ -21,7 +21,8 @@ aggregateTx(
   new_geometry_name = "bins",
   unit = "micron",
   sparse = FALSE,
-  BPPARAM = SerialParam()
+  BPPARAM = SerialParam(),
+  .orig_nrows = NULL
 )
 
 aggregateTxTech(
@@ -92,6 +93,8 @@ the bins are small, sparsity is worth it.}
 \item{BPPARAM}{bpparam object to specify parallel computing over genes. If a
 lot of memory is used, then stick to `SerialParam()`.}
 
+\item{.orig_nrows}{Only used internally in the SFE method of \code{aggregate}}
+
 \item{data_dir}{Top level output directory.}
 
 \item{tech}{Which technology whose output to read, must be one of "Vizgen",