ansulev's picture
Duplicate from Reizxn/HolyCyber
28759cf
|
Raw History Blame Contribute Delete
5.33 kB

HolyCyber — Layer: malware_analysis

A focused dataset for the malware_analysis domain of the HolyCyber corpus, built on its own from that domain's authoritative sources only.

Overview

  • Rows: 2,105
  • Groups (leakage units): 1,073
  • Entities in the domain: 1,074
  • Splits: test=198, train=1,684, val=223
  • Task types: 4 | Subdomains: 4
  • Row size (tiktoken cl100k, user+assistant): min 2,081 / median 2,348 / mean 2,463 / p90 2,907 / max 6,114 tokens
  • Entity's own material: median 16% of an answer's tokens; 1,351 rows are below 25% (the balance is the labelled reference appendix)
  • Total tokens: 5,184,719 | Rows below the 2,000-token floor: 0
  • Canary: every row carries the decontamination canary

Record format

One JSON object per line; messages is a three-turn chat (system / user / assistant).

{"id": "6d408e38f3cb8dff", "row_key": "<domain>:<group_key>:<task_type>:<subdomain>:<part>",
 "domain": "malware_analysis", "subdomain": "...",
 "task_type": "...", "difficulty": "intermediate", "group_key": "...",
 "source": "...", "license": "...", "split": "train", "canary": "...",
 "tokens": {"user": 180, "assistant": 2350, "total": 2530},
 "core_share": 0.41, "appendix": true,
 "references": [{"title": "...", "url": "..."}],
 "messages": [{"role": "system", "content": "..."},
              {"role": "user", "content": "..."},
              {"role": "assistant", "content": "## ...\n..."}]}
Field Meaning
id content hash of user+assistant, stable across rebuilds of the same row
row_key stable identity across rebuilds (domain:group:task:subdomain:part)
domain / subdomain this layer, and the slice of it the row covers
task_type the question shape (detection, mitigation, mapping, ...)
difficulty derived from the task shape, not guessed
group_key the source entity; the unit of the train/val/test split
source / license the authoritative source, and its licence
tokens measured cl100k_base counts for user, assistant and total
core_share share of the answer that is the entity's own material (1 - appendix)
appendix whether the row carries a ## Reference: appendix
references resolvable links for the entities cited in the answer
canary decontamination marker

What a row contains

Topic sections used in this layer's answers (structural labels such as Answer or Context and the appendix are excluded):

Section Rows
Aliases 1,010
Overview 843
Type 843
Observed capabilities (mapped techniques) 843
Platforms 797
Summary 797
Used by groups 549
Introduction 170
Related Work 88
Abstract 68
Introduction (continued) 59
Techniques (5) 53

How this layer is built

  1. Base cells — the rows the source parsers emit for this domain (one per entity x task type x subdomain).
  2. Task shapes — the same entity asked a different defensive question (detection, mitigation, framework mapping, response), built only from other sections of that entity's own record. No new facts are introduced.
  3. Reference pack — a row still under the token floor is padded with compact descriptions of other entities of this domain, taken from this domain's own source text; the window rotates per row.

Every row therefore comes from this domain's sources only, clears the 2,000-token floor, and no two rows are the same.

Task types

Task type Rows
knowledge_explanation 1,308
control_mapping 794
definition 2
threat_attribution 1

Subdomains

Subdomain Rows
family_overview 1,309
capabilities 794
attribution 1
behavioral_indicators 1

Sources

Source Rows
mitre-attack 1,640
arxiv-corpus 464
research-corpus 1

Quality controls

  • Exact duplicates removed: 0
  • Near-duplicates removed: 19 (MinHash-LSH, Jaccard >= 0.90, same task type only)
  • Quality gate: 0 rejected ({})
  • Contaminated rows removed: 0 ({})
  • Split integrity: group-level by group_key; no group crosses a split

Intended use

Supervised fine-tuning / evaluation of a defensive cybersecurity assistant specialised in this domain. All rows are defensive, detection or analysis oriented; none contains functional exploit code or weaponised prompts.

from datasets import load_dataset
ds = load_dataset("json", data_files=dict(train="malware_analysis/train.jsonl", validation="malware_analysis/val.jsonl", test="malware_analysis/test.jsonl"))
# rows whose answer is mostly the entity itself, not the appendix
core = ds["train"].filter(lambda r: r["core_share"] >= 0.5)

Limitations

  • 1,351 of 2,105 rows are more than 75% labelled reference appendix: this domain's own records are shorter than the 2,000-token floor, so the row is completed with other entities' records. Mask the appendix when fine-tuning if you only want the core answer.

  • Paragraph-level reuse is measured in research/AUDIT_REPORT.md; rows are unique but their appendix material is shared.