why does the llm continue to do back channeling, e...
# ❓・ai-evy-supportr
rosij
03/11/2026, 6:38 PM@AI Evy why does the llm continue to do back channeling, even when it's turned off? The agent still adds in um, like, etc and I don't want it to
a
AI Evy
03/11/2026, 6:38 PMThanks for tagging me @rosij. I'll look through relevant sources and get back to you shortly.
AI Evy
03/11/2026, 6:38 PM### Answer
Based on recent community reports, this appears to be a known issue that emerged around November 11th, 2025 [(1) ]() . Users reported a sudden increase in "Um"s being injected into agent responses, even with backchanneling turned off . The issue occurred uniformly across multiple agents without any configuration changes .
**Solution**: Prompt your agent to never use filler words . Add explicit instructions in your agent's prompt like "Never use filler words such as 'um', 'like', 'uh', etc." .
One user confirmed this approach resolved the issue . While the backchanneling setting should prevent this behavior, explicitly instructing the LLM appears to be the current workaround .
If you continue experiencing issues after adding this prompt instruction, I recommend:
* Visit [Contact Sales]()
* Ask questions on [Discord]()
* Visit
AI Evy
03/11/2026, 6:38 PM
To learn more, see the sources I considered:
[[1] Huge increase backchanneling "Um"s](https://discord.com/channels/1186424376300679209/1438582091905499239)
AI Evy
03/11/2026, 6:38 PMFeel free to tag @AI Evy with additional questions. I also appreciate feedback, it helps me improve.