Topic

Activation Steering

1 post about Activation Steering. Every summary links to the original.

See the idea

Activation Steering explainers

Loading Activation Steering: One Dial, Two Costs image
What is it?Activation Steering Evaluation Fine-Tuning LLMs Prompt Engineering
Activation Steering: One Dial, Two Costs
Activation steering edits a model's internal activations to add or remove a concept at inference time. This explainer maps its one control knob, its two costs, and when it beats prompting.

Paid picks we use ourselves. Links carry a referral code — a subscription through one earns us a commission at no extra cost to you.

Newest first

Apple
On the Effectiveness-Fluency Trade-Off in LLM Conditioning: A Systematic Study
Controlling the output of Large Language Models (LLMs) is a central challenge for their reliable deployment, yet current conditioning methods are often evaluated narrowly on effectiveness without considering generation quality. This study systematically investigates a range of conditioning methods in both concept injection and removal scenarios, including efficient steering, activation steering, prompting, and supervised fine-tuning. The authors find that efficient steering methods frequently achieve their conditioning goal at a steep cost to fluency, degrading generation quality. They identify a critical previously overlooked interaction: activation steering methods are far less effective on instruction-tuned models than on their base counterparts. Simple prompting and supervised fine-tuning are viable options for concept injection but are not as good at concept removal, while cheaply computed textual metrics highly correlate with costly LLM-as-judge scores, providing practical insights for evaluation.

The best engineering reads, once a week.

Topic

1 post about Activation Steering. Every summary links to the original.

See the idea

Loading Activation Steering: One Dial, Two Costs image
What is it?Activation Steering Evaluation Fine-Tuning LLMs Prompt Engineering
Activation Steering: One Dial, Two Costs
Activation steering edits a model's internal activations to add or remove a concept at inference time. This explainer maps its one control knob, its two costs, and when it beats prompting.

Paid picks we use ourselves. Links carry a referral code — a subscription through one earns us a commission at no extra cost to you.

Newest first

Apple
On the Effectiveness-Fluency Trade-Off in LLM Conditioning: A Systematic Study
Controlling the output of Large Language Models (LLMs) is a central challenge for their reliable deployment, yet current conditioning methods are often evaluated narrowly on effectiveness without considering generation quality. This study systematically investigates a range of conditioning methods in both concept injection and removal scenarios, including efficient steering, activation steering, prompting, and supervised fine-tuning. The authors find that efficient steering methods frequently achieve their conditioning goal at a steep cost to fluency, degrading generation quality. They identify a critical previously overlooked interaction: activation steering methods are far less effective on instruction-tuned models than on their base counterparts. Simple prompting and supervised fine-tuning are viable options for concept injection but are not as good at concept removal, while cheaply computed textual metrics highly correlate with costly LLM-as-judge scores, providing practical insights for evaluation.

The best engineering reads, once a week.