Repository navigation
fix: 入口画像准入时补足保留强度 - #117
Merged
Merged
Conversation
leafliber
added a commit
that referenced
this pull request
Oct 17, 2026
This was referenced Oct 17, 2026
Merged
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
入口画像话题/说法在候选阶段可能已经衰减到二十多;原实现准入只将 importance 提为 50,未补足保留强度,导致新准入画像仍很快遗忘。本 PR 在准入时原子补足到 importance 50 对应初值 50,并从准入重新开始普通衰减周期。
来源:按 EF-19 从 PR #116 的 be7f8fe 仅拆出准入保留强度部分。基于 main
134b21acc385fcccd00f9d7db34ed9653d5ab3b9新建独立分支;#116 保持开放,未改动其工作区、分支或说明。实现限于
memory_ops.write_entry_profile_memory:adjust_retention的原子下限round(30+50*.4)=50,并将decay_visits归零;更高保留强度保留。admission_retention(before/after/decay_visits);CAS、人工保护和来源回执保持 main 行为。保留强度自身不新增修订号,也不改证据确认时间。唯一适配是将原提交共用的类别查询移入
admit分支,以避免带入说法回执修改。main 的准入状态与写事务流程无需调整。未移植说法计数/变体、_inc_saying_key、证据选择、单次细节检查、任何提示词或默认值,也未移植 EF-16/17/18 报告。测试文件:新增
tests/test_entry_profile_admit_retention.py,从来源提交取 7 个保留强度案例,另补同正文实际复查、直接重放、规则类排除,共 10 项;审计、确认时间和修订号也有断言。tests/test_entry_profile_incremental.py仅取来源提交的一行预期值 47→50。先红后绿:未改产品代码时 5 failed/5 passed;首次定向回归唯一失败是上述旧预期值;取来对应断言后 97 passed。
学习对照采用仓库
evals/compare_learning_requests.py的外部副本,仅补齐公开 v5,未修改仓库运行器。对照全部使用假模型,不读取真实模型配置、不作语义质量判定。两版本顺序运行,均完成后清理本 worktree 的.pytest-tmp,.venv保持 3.13.13。Head:
c274fd5ace66c98e6cd18c3ddb07c888dfd55d86。完整验证材料与报告在/Users/cassia/Local/Code/iris-eval-artifacts/EF-19-20261017/。本轮不跑真实模型;不合并,待规划者审查。