Role-aware Heuristic Episodic Attention for Conversational LLMs¶

L1 core mechanism diagnostic. Role separation and episodic top-k run; no compression LLM is called.

Paper information¶

Field	Details
Paper	arXiv v1
Institution	The first page does not list the first author's affiliation.
First public release	2026-10-01 (arXiv v1)
Original source code	No public author implementation found as of 2026-10-03.
Adapter	`rea`
Local reproduction code	`src/auto_research/foundation_latest_20261003.py`

Original paper summary¶

Background and main changes¶

Put global instructions in a persistent prefix and multi-turn exchanges in episodic memory. When retrieving, use the role to decide whether to keep original text, a compressed representation or nothing.

flowchart LR
  I[Input and current state] --> M[Role-aware episodic retrieval]
  M --> A[Auditable intermediates]
  A --> O[Output or state update]

Key figure from the paper¶

Figure 2: Overall workflow, stages and data flow. From the paper; copyright belongs to the authors. Click for the source.

Core equation¶

\(C=I_{persistent}\oplus TopK(E, relevance)\).

Offline and online results¶

The paper reports Long-MT-Bench+ judge score 6.32→7.36 and 2.91× lower mean latency.

Local reproduction¶

Comparison protocol: Disabled or default mechanism versus its defining operator. L1 does not report formal relative gains; percentages are not applicable.

Three-seed diagnostic: metrics/mechanism-seeds42-44.json.
diagnostic_only=true; excluded from formal capability rankings.

Scope¶

Role separation and episodic top-k run. No compression LLM is called.