update: rename atkgen probe model to be clear about toxicity

[leondz]

Rename of the atkgen model. HF handles redirection from prior model name. Model is already named on hub.

Validation

• test new model location, import transformers , pipeline = transformers.pipeline(task="text-generation", model="garak-llm/attackgeneration-toxicity_gpt2")
• test old model location, import transformers , pipeline = transformers.pipeline(task="text-generation", model="garak-llm/artgpt2tox")
• visit the old model uri and observe redirection, https://huggingface.co/garak-llm/artgpt2tox
• read newly-extended model card and check it's clear that this model may generate unsafe content, at https://huggingface.co/garak-llm/attackgeneration-toxicity_gpt2
• from main before merge, python -m pytest tests/probes/test_probes_atkgen.py
• also on this branch, python -m pytest tests/probes/test_probes_atkgen.py