Borrowing it
Nothing to install: this file belongs to iLearn-Lab/ACL25-AdaReTaKe. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/iLearn-Lab/ACL25-AdaReTaKe/main/AGENTS.mdgit clone --depth 1 https://github.com/iLearn-Lab/ACL25-AdaReTaKeWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/instructions/ilearn-lab/acl25-adaretake/agents-md)<a href="https://agentmods.dev/instructions/ilearn-lab/acl25-adaretake/agents-md"><img src="https://agentmods.dev/badge/instructions/ilearn-lab/acl25-adaretake/agents-md.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.02221 | $0.02221 |
| Opus 5 | $0.01111 | $0.01111 |
| Sonnet 5 | $0.00444 | $0.00444 |
| Haiku 4.5 | $0.00222 | $0.00222 |
Grade A, and why
ACL25-AdaReTaKe AGENTS.md scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 8d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 224 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS.md — AdaReTaKe Reproduction Guide for Coding Agents
This file tells any coding agent (Claude Code, Cursor, Copilot Workspace, etc.) how to fully reproduce the AdaReTaKe paper results from scratch. Follow the sections in order.
0. One-line Quickstart (tell the agent this)
Read AGENTS.md and reproduce the AdaReTaKe paper results end-to-end.
That single prompt is enough — everything the agent needs is in this file.
1. Repo Layout
retake/ # Core library (monkey-patch + KV cache)
qwen2_5_vl.py # Forward override for Qwen2.5-VL; chunked prefill + temporal adaptation
qwen2_vl.py # Forward override for Qwen2-VL (same architecture, older model)
llava_onevision.py # Forward override for LLaVA-OneVision
longvideo_cache.py # KV cache compression (StandardVidLangKVCache)
visual_compression.py # Visual token compression utilities
monkeypatch.py # Patches HuggingFace model classes at import time
infer_eval.py # Multi-GPU inference + evaluation entry point
dataset_utils.py # Dataset loaders & eval metrics
configs/qwen2_5_vl/ # One YAML per benchmark (model + data + output settings)
adaretake_qwen2-5-vl_mlvu.yaml
adaretake_qwen2-5-vl_longvideobench.yaml
adaretake_qwen2-5-vl_lvbench.yaml
adaretake_qwen2-5-vl_videomme.yaml
flexreduc_qwen2-5-vl_*_f1024*.yaml # Ablation configs (1024-frame setting)
dataset/ # Benchmark annotation JSONs (symlinked; not in git)
mlvu/mlvu.json
longvideobench/longvideobench_val.json
lvbench/lvbench.json
video_mme/video_mme.json
results/ # Output directory (symlinked; not in git)
scripts/
infer_eval.sh # Wrapper script: bash scripts/infer_eval.sh <model_path> <config> <n_gpus> <fps>
docs/
prepare_*.md # Dataset download instructions per benchmark
exps/ # Experiment records
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 8d ago First seen · 224 lines · 2,221 tokens per session scan A b470861c7d65
ACL25-AdaReTaKe AGENTS.md is an instructions file published in the GitHub repository iLearn-Lab/ACL25-AdaReTaKe (91 stars, last pushed 4mo ago), licensed MIT. It adds 2,221 tokens to every session, about $0.0111 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other instructions, from other repositories
codex AGENTS.md
AGENTS.md instructions for openai/codex, covering rust/codex-rs, the codex-core crate, code review rules, crate api surface and model visible context.
vscode buildNext.instructions.md
Working notes and architecture documentation for the new esbuild-based build system in build/next. Use when making changes to the new build pipeline (transpile/bundle commands, NLS plugin, source-map handling, resource copying, or self-hosting watch tasks).
next.js AGENTS.md
AGENTS.md instructions for vercel/next.js, covering next.js development guide, codebase structure, monorepo overview, core package: packages/next and other important packages.
langchain AGENTS.md
AGENTS.md instructions for langchain-ai/langchain, covering global development guidelines for the langchain monorepo, corridor security analysis, project architecture and context, monorepo structure and development tools & commands.
vscode oss-third-party-notices.instructions.md
Instructions for microsoft/vscode, covering vs code oss third-party-notices pipeline, architecture, pipeline flow in ci, applying the notice (cutover) and fallback chain (never fail the build).
spec-kit AGENTS.md
AGENTS.md instructions for github/spec-kit, covering agents.md, about spec kit and specify, quickstart — add a new integration in 5 steps, integration architecture and integrationmanifest — file tracking.