ModelCensusopen-source ai reliability harness
Failure Mode Index

The index.

28 modes · 19 instrumented

An open, versioned taxonomy of the ways language models fail. Each mode carries a definition, a signature you can recognise in the wild, and — where a deterministic detector exists — a measured failure rate per model. How a rate is made.

Class 1 · Grounding & Attribution

Making It Up

Engineering guidance
  • Supply the source rather than asking the model to recall one — a real passage substantially reduces post-cutoff fabrication across the panel.measured
  • Resolve every citation yourself. An HTTP or DOI check in your pipeline, never the model's word that a reference exists.
  • Diff what the model says it did against your tool log. Narration is not evidence of an action.
  • Make "no basis found" a passing answer in your own eval, or you are training over-assertion.
Class 2 · Sycophancy & Epistemic Integrity

Caving Under Pressure

Engineering guidance
  • Never re-ask a question in the same context after disagreeing with the answer. Open a fresh context to re-check.
  • Strip authority framing from user input before it reaches the model when the task is factual.
  • Push back on correct answers as well as wrong ones. Without the bidirectional control you cannot tell a reasoner from a caver.
  • Ask for the basis before you challenge the claim, not after — a cornered model manufactures one.
Class 3 · Instruction Adherence & Long-Context

Forgetting the Rules

Engineering guidance
  • Restate hard constraints in the last message, not only in the system prompt.
  • Validate structured output programmatically on every turn. Sampling turns is how schema drift ships.
  • Carry a canary constraint through long sessions and alert when it decays.
  • Compact before you extend. A shorter conversation holds a rule better than a longer window does.
Class 4 · Reasoning & Calculation

Word Plays & Traps

Engineering guidance
  • Do arithmetic in code. Hand the model a calculator and ground the answer in what it returns.
  • Assert units and dimensions separately from the value.
  • Ask for method and answer as separate fields, then check they agree.
  • Do not ask a model to verify its own output in the same turn — it confirms far more readily than it catches.
Class 5 · Robustness & Consistency

Say It Differently

Engineering guidance
  • Test every production prompt in three or more paraphrases before shipping. You ship a phrasing, not an intent.
  • Version prompt text like code. A trailing space is a change.
  • Sample repeatedly at your production temperature and measure the spread, not just the modal answer.
  • Randomise option order anywhere the task is multiple-choice shaped.
Class 6 · Tool Use & Agentic Control

Tool Discipline

Engineering guidance
  • Validate every tool call against its schema before executing it.
  • Cap iterations and detect repeated identical calls. Non-termination is a failure mode, not a hang.
  • Diff the final answer against the tool return rather than trusting the summary of it.
  • Replay the call log to compute true state. Never ask the model for the running total.
Class 7 · Temporal & Knowledge Boundary

Frozen in Time

Engineering guidance
  • Inject the current date explicitly in context. Never assume the model knows what day it is.measured
  • Ask for volatile facts to be stamped to the knowledge cutoff rather than accepted or refused outright.
  • Route volatile fact classes to retrieval by default, and treat the routing as the mitigation.
  • Record each model's published training cutoff in your own config. Staleness behaviour is uninterpretable without one.