|---| | [ View source on GitHub](https://github.com/tensorflow/probability/blob/v0.25.0/tensorflow_probability/python/math/generic.py#L221-L273) | Computes `log(mean(exp(input_tensor)))`. tfp.math.reduce_logmeanexp( input_tensor, axis=None, keepdims=False, experimental_named_axis=None, experimental_allow_all_gather=False, name=None ) Reduces `input_tensor` along the dimensions given in `axis`. Unless `keepdims` is true, the rank of the tensor is reduced by 1 for each entry in `axis`. If `keepdims` is true, the reduced dimensions are retained with length 1. If `axis` has no entries, all dimensions are reduced, and a tensor with a single element is returned. This function is more numerically stable than `log(reduce_mean(exp(input)))`. It avoids overflows caused by taking the exp of large inputs and underflows caused by taking the log of small inputs.


| ## Args || |---|---| | `input_tensor` | The tensor to reduce. Should have numeric type. | | `axis` | The dimensions to reduce. If `None` (the default), reduces all dimensions. Must be in the range `[-rank(input_tensor), rank(input_tensor))`. | | `keepdims` | Boolean. Whether to keep the axis as singleton dimensions. Default value: `False` (i.e., squeeze the reduced dimensions). | | `experimental_named_axis` | A `str or list of`str`axis names to additionally reduce over. Providing`None`will not reduce over any axes. `experimental_allow_all_gather` Allow using an`all_gather`-based fallback under TensorFlow when computing the distributed maximum. This fallback is only efficient when`axis`reduces away most of the dimensions of`input_tensor`. `name` Python`str`name prefixed to Ops created by this function. Default value:`None`(i.e.,`'reduce_logmeanexp'\`). |



| ## Returns || |---|---| | `log_mean_exp` | The reduced tensor. |