Anthropic launches Conceptual Reasoning Index to measure AI's philosophical thinking
Anthropic: Introducing The Conceptual Reasoning Index
Anthropic and Redwood Research introduce the Conceptual Reasoning Index (CRI), a suite of three benchmarks—LMCA, ACCoRD, and DTBench—to evaluate AI's ability to reason about conceptual questions where empirical feedback is limited. Initial results show top models score 73.6 out of an estimated ceiling of 91, with progress roughly linear and no signs of flattening.
A core hope for managing AI risks is that AIs will help us understand our situation, plan for what lies ahead, and develop risk mitigations.
- lanyard-textile
Ah, yes -- A closed source benchmark that Anthropic paid for that Anthropic ranked highest.
0/10
- andsoitis
Opening line:
> A core hope for managing AI risks is that AIs will help us understand our situation
Gonna stop you right there and ask that you think deeply about that premise.
- xlayn
We have totally come out with this idea of this index that will allow us to create policies to ensure only the best and safest AIs are used by the public....
Absolutely no conflict of interest, no lobbying here
And no this is not related to those chinesse models... it's not the same as HD vs honda thing...
Trust us, this is the same kind of amazing thing as boeing doing their own certifications and inspections!
-- First reply: those dumb models, who use them, they are good only for adding 1 + 1
-- another: They will do the same so who cares...
-- The valve guy is a dick, and has a monopoly...
- onomojo
We invented a new benchmark and look we're at the top. Everyone else sucks compared to us. Especially those dirty open models.
- eutropia
This is a modest start on an important direction for AI Alignment work; which is, as the authors observe, commonly comprised of tasks which are not readily empirically verifiable and not easily mathematically modeled - so it's hard to get at with normal RL techniques.
I find the ACCoRD benchmark the most interesting, because you could theoretically scale it up from the baseline mode of testing two instances of the same model for their `P(A) ≥ P(A&B)` respectively, you could do `P(A)≥ P(A&B) && P(A) ≥ P(A&C) && P(A&B) ≥ P(A&B&C) && P(A&C) ≥ P(A&B&C) ...` etc
i.e. a swarm of model instances could be collectively measured for consistency for even more confidence, right?
At any rate, even the basic idea of measuring a model for consistency in beliefs improves our ability to bound the amount of trust we can put on it with introspection methods.