FA-98596 / KV cache position mapping / Open access
Prefix reuse reports the common token prefix length · case 01
Reuse stops at the first differing token but reports the final sequence length.
ROOT CAUSE
The comparison does not track the last matching boundary.
THE FAILURE
The comparison does not track the last matching boundary.
Unsuccessful approach: Comparing membership instead of aligned token equality overestimates reuse for reordered tokens.
Case contract
Return the length of the longest shared prefix of two token sequences.
Why this case matters
A small in-memory model makes the cache address contract explicit for a serving component.
1 / The failure
Exit 1"""Failure Map reference implementation. Python standard library only."""
import json
N = 1
observations = []
def solve(cached_tokens, requested_tokens):
return min(len(cached_tokens),len(requested_tokens))
def check(label, actual, expected):
observations.append({"check": label, "actual": actual, "expected": expected, "passed": actual == expected})
check('full prefix',solve(['a']*N+['b'],['a']*N+['c']),N)
check('first mismatch',solve(['x','a'],['y','a']),0)
check('short request',solve(['a','b','c'],['a','b']),2)
check('empty cached',solve([],['a']),0)
check('empty request',solve(['a'],[]),0)
check('reordered tokens',solve(['a','b'],['b','a']),0)
print(json.dumps({"observations": observations, "passed": all(x["passed"] for x in observations)}, ensure_ascii=False))
raise SystemExit(0 if all(x["passed"] for x in observations) else 1)
| Boundary fixture | Actual | Expected | Outcome |
|---|---|---|---|
| full prefix | 2 | 1 | Failed |
| first mismatch | 2 | 0 | Failed |
| short request | 2 | 2 | Passed |
| empty cached | 0 | 0 | Passed |
| empty request | 0 | 0 | Passed |
| reordered tokens | 2 | 0 | Failed |
SHA-256 / 7133e21b105f74d510af1e056b35c25b375e54cf203db65d01d7174f888288b7
2 / The unsuccessful fix
Exit 1"""Failure Map reference implementation. Python standard library only."""
import json
N = 1
observations = []
def solve(cached_tokens, requested_tokens):
return sum(t in requested_tokens for t in cached_tokens)
def check(label, actual, expected):
observations.append({"check": label, "actual": actual, "expected": expected, "passed": actual == expected})
check('full prefix',solve(['a']*N+['b'],['a']*N+['c']),N)
check('first mismatch',solve(['x','a'],['y','a']),0)
check('short request',solve(['a','b','c'],['a','b']),2)
check('empty cached',solve([],['a']),0)
check('empty request',solve(['a'],[]),0)
check('reordered tokens',solve(['a','b'],['b','a']),0)
print(json.dumps({"observations": observations, "passed": all(x["passed"] for x in observations)}, ensure_ascii=False))
raise SystemExit(0 if all(x["passed"] for x in observations) else 1)
| Boundary fixture | Actual | Expected | Outcome |
|---|---|---|---|
| full prefix | 1 | 1 | Passed |
| first mismatch | 1 | 0 | Failed |
| short request | 2 | 2 | Passed |
| empty cached | 0 | 0 | Passed |
| empty request | 0 | 0 | Passed |
| reordered tokens | 2 | 0 | Failed |
SHA-256 / 634f6c3c16b808189b473db3a457f7c033624ff27aeb07dd13ab844706744825
HELD IN THE MEMBER ARCHIVE
The verified repair and its recorded checks are member-only.
This mechanism has 6 recorded checks per implementation. The open-access tier publishes the failure and the unsuccessful fix; the repaired source that passes every check, and the observations that prove it, are available to members.
Every case sharing this mechanism uses the same contract and the same repair, so this one record is held back for all of them.
Member access is invitation-based. Sign in with your invited account to inspect the repair.
Sign in to the archive ↗Verification & scope
These toy integer maps omit tensor layout, concurrency, device memory, and model-specific cache formats. This reproducer isolates one failure mechanism. Results cover the supplied fixtures. Variants within a family share a test contract and should remain grouped when constructing evaluation splits. Related mechanisms with a shared evaluation_group must also remain together; these controlled models are not independent production incidents.
Observations recorded using Python 3.12.14 at 2026-09-30T16:48:14.544796+00:00.
Case digest / 0dd48b74e1de07e0cff18a7d0dac0cbfb86c843445008b7d86922a87344ad0e6