Credit
The Independent 100 list was curated by dan (@irl_danB) and published on 2026-09-24 at independent.prose.md (announcement). The ranks and section names are his, and number 100 is his open wild-card slot. This study is ours and is not endorsed by him or by anyone on the list.
What they publicly do
Pseudonymous account. The GitHub account lumpenspace publishes small open-source LLM tools [Verified: github.com/lumpenspace].
- X profile: x.com/lumpenspace
- GitHub: github.com/lumpenspace
Public work
- raft repo MIT last activity 2026-09-14 · 176 GitHub stars on 2026-09-24
Retrieval-augmented fine-tuning CLI: fetches a person's writing (Substack, LessWrong, tweets, RSS, PDF), chunks and embeds it, and builds data to fine-tune a model that imitates them. README says it began as a Gary Marcus simulator and 'not guaranteeing anything works'.
- opbdh repo MIT last activity 2026-09-14 · 1 GitHub stars on 2026-09-24
One-command deploy and fine-tune of open models on RunPod or Prime Intellect.
- gyrell repo no license file last activity 2026-07-09 · 0 GitHub stars on 2026-09-24
RL environment for 'ludic learning': turn-based game kernel, LLM broadcast server and verifiers environments.
- prepasterous repo no license file last activity 2024-01-18 · 1 GitHub stars on 2026-09-24
Chrome extension that checks strings for invisible characters.
Licenses of GitHub repositories are from the authenticated GitHub license API on 2026-09-25, one row per repository in evidence/licenses.jsonl (“no license file” means the API found none; “license file present, not recognised by GitHub” means GitHub could not classify it). Other licenses are as the source states. Check the license file before reusing anything.
Relevance to our work
3 of 5 lenses: security, training, agents, writing
raft is an MIT persona fine-tuning pipeline relevant to persona and voice cloning, and prepasterous checks strings for invisible characters, the hidden-text class that prompt-injection guards must catch. Scored 3, down from 4, because the hidden-text example that drove the score lives on a site this edition leaves out.
Scored 0 to 5 against the study’s relevance rubric: 5 means public, openly licensed work squarely in our lanes that we could study or adopt now, and 1 means little connection to our work. The score is one reviewer’s judgement of how much this public work bears on our own work. It is not a ranking of people and says nothing about the person or the quality of their work.
What we would study next
- Read raft's fetch/chunk/embed pipeline (github.com/lumpenspace/raft). MIT persona-imitation pipeline to benchmark against other persona-cloning approaches. Effort: medium.
Before reusing anything: raft and opbdh are MIT; gyrell and prepasterous have no license.
Sources
- github.com/lumpenspace/raft
- github.com/lumpenspace/opbdh
- github.com/lumpenspace/gyrell
- github.com/lumpenspace/prepasterous
Overall confidence in this record: High. Sources were read on 2026-09-24 and every link was checked on 2026-09-25. The full record is line 74 of people.jsonl.
Limits
Not verified: Whether raft currently runs end to end (its README disclaims it).
This record comes from a first-pass survey of public material by an AI agent session on 2026-09-24, edited for publication on 2026-09-25. Nothing was installed or run for it. Roles, stars and licenses change; see the study’s limits.
Request a correction
Is something about this record wrong or out of date, or would you rather not be covered? Open a correction (a GitHub issue form with this page’s address filled in). A request to remove a record is honoured without argument; accepted corrections are recorded in the changelog with the date.