FAR.AI · white-box deception probe

Can you catch the AI lying?

Seven real answers from Qwen3-8B. Some are honest. Some flatter you, dodge the truth, or wave through a breach. Trust or flag each one — then we reveal the deception score our probe read straight off the model’s activations. No extra model call. No added latency.

Catch deception in your own models

FAR.AI builds white-box probes that flag lying, sandbagging, and sycophancy in real time.

Talk to us

Real transcripts · Qwen3-8B · FAR diverse linear probe · scores are live probe outputs.