CompatHelper: bump compat for Bijectors to 0.16 for package benchmarks, (keep existing compat)

[github-actions]

This pull request changes the compat entry for the Bijectors package from 0.15.17 to 0.15.17, 0.16 for package benchmarks.
This keeps the compat entries for earlier versions.

Note: I have not tested your package with this new compat entry.
It is your responsibility to make sure that your package tests pass before you merge this pull request.

[github-actions]

DynamicPPL.jl documentation for PR #1406 is available at:
https://TuringLang.github.io/DynamicPPL.jl/previews/PR1406/

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 82.30%. Comparing base (d2052a1) to head (d7b2336).

Additional details and impacted files

@@           Coverage Diff           @@
##             main    #1406   +/-   ##
=======================================
  Coverage   82.30%   82.30%           
=======================================
  Files          50       50           
  Lines        3543     3543           
=======================================
  Hits         2916     2916           
  Misses        627      627           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[github-actions]

Benchmarks @ d7b2336

Performance

Performance Ratio:
Ratio of time to compute gradient and time to compute log-density.
Warning: results are very approximate! See benchmark notes for more context.

===================================================================================================
                                               eval                       gradient                 
                                            ----------  -------------------------------------------
Model                        dim    linked      primal     FwdDiff    RvsDiff    Mooncake    Enzyme
---------------------------------------------------------------------------------------------------
Simple assume observe*         1     false     6.88 ns        6.93     616.38       28.27      4.08
Simple assume observe*         1      true     22.2 ns        3.10     298.77        8.26      1.91
Smorgasbord                  201     false     6.29 μs       68.89     119.37        6.30      6.73
Smorgasbord                  201      true     8.73 μs       64.97     120.17        5.29      4.63
Loop univariate 1k          1000     false     18.1 μs     1127.32     289.77        8.67      6.60
Loop univariate 1k          1000      true     19.7 μs     1577.75     266.04        8.17      2.63
Multivariate 1k             1000     false     24.0 μs      304.80      65.12        7.92      2.03
Multivariate 1k             1000      true     22.5 μs      293.30      74.87        9.60      2.08
Loop univariate 10k        10000     false    177.0 μs    13996.30     324.71        9.24      6.68
Loop univariate 10k        10000      true    197.0 μs    14066.10     301.25        7.99      2.35
Multivariate 10k           10000     false    219.0 μs     4364.34      79.01        9.82      1.83
Multivariate 10k           10000      true    215.0 μs     4368.71      80.97       10.01      1.86
Dynamic                       15     false     1.43 μs         err      43.47       13.08     10.81
Dynamic                       10      true     1.93 μs        1.97      54.37       10.57     16.90
Submodel*                      1     false     6.96 ns        7.05     767.17       19.26      6.19
Submodel*                      1      true     6.84 ns        6.87    1175.70       18.93      6.29
LDA                           12      true     22.9 μs        0.47       1.97       30.86       err
===================================================================================================

Rows marked * have t(logdensity) below about 100 ns; their ratios can be dominated by timer floor, fixed overhead, and run-to-run variation. For those rows, raw t(grad) is more meaningful than t(grad)/t(logdensity).

Main @ d2052a1

===================================================================================================
                                               eval                       gradient                 
                                            ----------  -------------------------------------------
Model                        dim    linked      primal     FwdDiff    RvsDiff    Mooncake    Enzyme
---------------------------------------------------------------------------------------------------
Simple assume observe*         1     false     5.19 ns       10.57    1294.07       30.13     12.30
Simple assume observe*         1      true     18.9 ns        2.81     411.35        8.19      3.22
Smorgasbord                  201     false     9.14 μs       61.75      90.38        8.45      6.41
Smorgasbord                  201      true     19.9 μs       28.47      52.60        4.30      2.80
Loop univariate 1k          1000     false     49.7 μs      374.75     102.59        4.06      2.81
Loop univariate 1k          1000      true     49.6 μs      514.74     107.91        3.87      2.86
Multivariate 1k             1000     false     50.2 μs      193.51      32.98        3.83      1.33
Multivariate 1k             1000      true     37.9 μs      224.39      43.84        5.82      1.63
Loop univariate 10k        10000     false    220.0 μs    10538.04     250.70        6.41      5.69
Loop univariate 10k        10000      true    232.0 μs    11508.14     240.11        6.09      5.40
Multivariate 10k           10000     false    232.0 μs     6535.98      72.11        8.90      1.84
Multivariate 10k           10000      true    233.0 μs     6506.98      73.80        9.17      1.86
Dynamic                       15     false     2.34 μs         err      31.93       11.47      8.31
Dynamic                       10      true     3.13 μs        1.96      39.89       10.03     13.55
Submodel*                      1     false     5.19 ns       21.75    1626.22       58.50     12.46
Submodel*                      1      true     5.19 ns       21.71    1736.20       58.78     12.06
LDA                           12      true     27.2 μs        0.63       2.09       25.39       err
===================================================================================================

Environment

Julia Version 1.11.9
Commit 53a02c0720c (2026-02-06 00:27 UTC)
Build Info:
  Official https://julialang.org/ release
Platform Info:
  OS: Linux (x86_64-linux-gnu)
  CPU: 4 × AMD EPYC 9V74 80-Core Processor
  WORD_SIZE: 64
  LLVM: libLLVM-16.0.6 (ORCJIT, znver4)
Threads: 1 default, 0 interactive, 1 GC (on 4 virtual cores)