Adding post instruction tokens for steering

[SevKod]

Introducing the ability to get activations and steering on the post-instruction tokens (as described in the appendix of https://arxiv.org/pdf/2406.11717 )

[codecov]

Codecov Report

❌ Patch coverage is 7.96020% with 185 lines in your changes missing coverage. Please review.
✅ Project coverage is 53.83%. Comparing base (75a2743) to head (3865a98).
⚠️ Report is 181 commits behind head on main.

Files with missing lines	Patch %	Lines
src/sdialog/interpretability/__init__.py	7.18%	168 Missing ⚠️
src/sdialog/agents.py	15.00%	17 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #103      +/-   ##
==========================================
+ Coverage   46.82%   53.83%   +7.01%     
==========================================
  Files          20       34      +14     
  Lines        4171     6133    +1962     
==========================================
+ Hits         1953     3302    +1349     
- Misses       2218     2831     +613     

Files with missing lines	Coverage Δ
src/sdialog/agents.py	50.28% <15.00%> (-1.79%)	⬇️
src/sdialog/interpretability/__init__.py	15.57% <7.18%> (-5.91%)	⬇️

... and 16 files with indirect coverage changes

🚀 New features to boost your workflow:

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

[sergioburdisso]

this block of 3 lines of code can be replaced simply by activation_index = self.token_index