While you can make a lot of progress in evals with tinkering and paying little attention to the literature, we found that various other papers have saved us many months of research effort. The Apollo Research evals team thus compiled a list of what we felt were important evals-related papers. We likely missed some relevant papers, and our recommendations reflect our personal opinions.

Our favorite papers

• Evaluating Frontier Models for Dangerous Capabilities (Phuong et al., 2024)
  • Contains detailed descriptions of multiple LM agent evals across four categories. Also explores new methodologies for estimating evals success probabilities. 
  • We think it is the best “all around” evals paper, i.e. giving the best understanding of what frontier LM agent evals look like
  • We tested the calibration of their new methodologies in practice in Hojmark et al., 2024, and found that they are not well-calibrated (disclosure: Apollo involvement).
• Observational Scaling Laws and [...]

---

Outline:

(00:29) Our favorite papers

(03:20) Other evals-related publications

(03:25) LM agents

(05:58) Benchmarks

(09:05) Science of evals

(12:02) Software

(13:03) Miscellaneous

(16:49) Related papers from other fields

(16:53) Red teaming

(18:13) Scalable oversight

(19:28) Scaling laws and emergent behaviors

(21:45) Science tutorials

(23:20) LLM capabilities

(25:15) LLM steering

(28:52) Fairness, bias, and accountability

(29:48) AI Governance

(32:31) Contributions

---

First published:

Source:

Narrated by TYPE III AUDIO.