The authors built Codetta around a shared public model that estimates the communication channel, a sampling procedure designed to preserve the sender’s normal output distribution, and an adaptive error-correcting code that compensates for decoding errors.
Agents can hide high-capacity messages without a pre-shared key
Codetta lets asymmetric language-model agents establish a secret key and communicate covertly while preserving the sender’s output distribution.
Top University
Qi Pang · Virginia Smith · Wenting Zheng
Carnegie Mellon University
Research Digest··2 min read
Pang, Smith and Zheng present a steganographic protocol for independently deployed agents that do not share the same model or an initial secret key.
Why this paper
From Carnegie Mellon University
In one line
Codetta enables high-capacity, undetectable steganographic collusion between independently deployed LLM agents without a pre-shared key.
What we could check
- ·No code link found
- ·No weights link found
- ·No dataset link found
- ·No compute details found
- ·No stated limitations found
- ·No benchmark numbers found
Observed from the paper text and links we have. Absence here means we did not find it, not that it does not exist.
§