LiveCodeBench crashes on iteration due to dataset type inconsistency

## Problem
LiveCodeBench variants can crash during execution due to:
1. Dataset not being converted to list format
2. Missing type validation for examples

## Reproduction Steps
```python
# This can fail depending on dataset implementation
from eval.chat_benchmarks.LiveCodeBench.eval_instruct import LiveCodeBench
benchmark = LiveCodeBench()
# Iteration may fail or behave unexpectedly
```