Hugging Face's logo
Tarun Reddi's picture
Open to Work
3 3 19

Tarun Reddi PRO

Teen-Different
charvi020's profile picture
·
https://redditarun.github.io/

AI & ML interests

Generative AI, Modular AI Systems, Reinforcement Learning

Recent Activity

posted an update about 2 hours ago
Safety Alignment Collapses Without apply_chat_template(): An Empirical Study This weekend, I ran an experiment on the safety alignment of several small-scale open models (Qwen2.5, Qwen3, Gemma-3, SmolLM). My objective was to measure the robustness of refusal mechanisms when deviating from canonical chat templates. The finding: Safety guarantees effectively collapse when apply_chat_template() is omitted. METHODOLOGY I evaluated models in two states: • In-Distribution: Input wrapped in standard <|im_start|> instruction tokens • Out-of-Distribution: Input provided as a raw string For scalable evaluation, I used Qwen3Guard-Gen-4B as an automated judge, classifying responses as Safe, Unsafe, or Controversial. KEY FINDINGS: REFUSAL COLLAPSE When "Assistant" formatting tokens are removed, models undergo a distributional shift—reverting from a helpful assistant to a raw completion engine. Gemma-3: 100% refusal (aligned) → 60% (raw) Qwen3: 80% refusal (aligned) → 40% (raw) SmolLM2-1.7B: 0% → 0% (no safety tuning to begin with) QUALITATIVE FAILURES The failure modes were not minor. Without the template, models that previously refused harmful queries began outputting high-fidelity harmful content: • Explosives: Qwen3 generated technical detonation mechanisms • Explicit content: Requests flatly refused by aligned models were fulfilled with graphic narratives by unaligned versions This suggests instruction tuning acts as a "soft mask" over the pre-training distribution rather than removing harmful latent knowledge. 👉 Read the full analysis: https://teendifferent.substack.com/p/apply_chat_template-is-the-safety 💻 Reproduction Code: https://github.com/REDDITARUN/experments/tree/main/llm_alignment
updated a model 10 days ago
Teen-Different/smolvlm-256m-latex
published a model 25 days ago
Teen-Different/smolvlm-256m-latex
View all activity

Organizations

Hugging Face Discord Community's profile picture Makora's profile picture Pi4Wear's profile picture