Poster
Benchmarking Sub-Billion-Parameter Models on Commodity Hardware
A poster summarising latency, memory and quality for twenty sub-billion-parameter models on a laptop CPU, a single-board computer and a mobile phone.
Reviewed by the event organisers
CONTINUUM Proceedings
Permanent, citable homes for conference papers.
Open-access proceedings for conferences, workshops and symposia. Organisers deposit their accepted papers; every paper and every volume receives a DOI.
- ✓ DOI for every paper and volume
- ✓ ISBN for each volume
- ✓ Organisers keep control of their review process
- Volumes
- 3
- Papers
- 16
- Downloads
- 3,944
Papers
4 resultsSearch & filters ↓
Full paperBest paper
Retry Budgets for Tool-Using Agents: When Is Trying Again Worth It?
Agents that retry failed tool calls recover from transient errors but can also compound mistakes and waste resources. We analyse 18,000 recorded agent runs and show that retries help for network and rate-limit errors but hurt for malformed arguments. We propose a retry budget that depends on the error class and show that it improves success by 6 points at equal cost.
Short paper
Detecting Silent Tool Failures from Output Distributions
Some tool failures return plausible but wrong output and raise no error. We detect such silent failures by comparing the distribution of tool outputs against a rolling reference and flagging shifts. On three deployed tools the detector finds 71% of injected faults with a 2% false-alarm rate.
Extended abstract
Perturbation Suites for Measuring Agent Reliability: Lessons from Three Deployments
We report experience building perturbation suites — rephrased instructions, reordered tool lists, noisy outputs — for three deployed agents, and summarise which perturbations revealed real failures and which only added cost.