fix: Bugs & improvements in explainers and explain scripts

[dilyabareeva]

• Extending evaluation: text_classification support for Representer Points wrapper, add kronfluence eval configs
• Bug fixes: mislabeling curve, train shuffle, module class method from_pretrained_base takes num_labels, CaptumSimilarity signature-order fix
• Clean-up: assert→typed exceptions, better eval scripts, dataset handling in base explainer, cleaner cache regeneration, HF-offline flags.
• Test updates.

[codecov]

Codecov Report

❌ Patch coverage is 96.24060% with 5 lines in your changes missing coverage. Please review.
✅ Project coverage is 96.14%. Comparing base (e9aa62a) to head (c178d32).
⚠️ Report is 29 commits behind head on main.

Files with missing lines	Patch %	Lines
quanda/benchmarks/base.py	83.33%	1 Missing ⚠️
quanda/explainers/wrappers/representer_points.py	92.85%	1 Missing ⚠️
quanda/metrics/heuristics/mixed_datasets.py	50.00%	1 Missing ⚠️
quanda/metrics/heuristics/model_randomization.py	85.71%	1 Missing ⚠️
...uanda/utils/datasets/transformed/label_flipping.py	50.00%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #284      +/-   ##
==========================================
- Coverage   96.19%   96.14%   -0.05%     
==========================================
  Files          63       64       +1     
  Lines        2917     3010      +93     
==========================================
+ Hits         2806     2894      +88     
- Misses        111      116       +5     

☔ 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.