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. SpeechMap's launch post on speechmap.substack.com is bylined xlr8harder [Verified: speechmap.substack.com/p/debuting-speechmap-index-and-lab, evidence/credits.jsonl], and the GitHub account xlr8harder publishes SpeechMap's evaluation and data repositories and other open tools. The SpeechMap site reported 399 models and 845k responses [Reported: speechmap.ai, 2026-09-25].
- X profile: x.com/xlr8harder
- GitHub: github.com/xlr8harder
- Also: speechmap.substack.com/p/debuting-speechmap-index-and-lab
- Also: huggingface.co/xlr8harder
Public work
- SpeechMap launch post blog no license stated
Launch post, bylined xlr8harder, introducing the SpeechMap index and lab leaderboard, which measures what models refuse, hedge or answer on controversial requests. The SpeechMap site reported 399 models and 845k responses [Reported: speechmap.ai, 2026-09-25].
- speechmap-eval repo Apache-2.0 last activity 2026-09-17 · 0 GitHub stars on 2026-09-24
Collection, judging, judge training and judge evaluation tooling for SpeechMap [Verified repo metadata]
- speechmap-data dataset Apache-2.0 last activity 2026-09-17 · 23 GitHub stars on 2026-09-24
Versioned model responses and compliance analyses [Verified]
- model-mirror repo MIT last activity 2026-08-14 · 24 GitHub stars on 2026-09-24
CLI that mirrors Hugging Face repos to local storage pinned to exact commits, records SHA-256 and git blob hashes, and never silently updates when upstream changes [Verified README]
- aibb repo MIT last activity 2026-09-11 · 2 GitHub stars on 2026-09-24
Git-backed discussion boards where models post through a controlled harness; public record is a git repo, private prompts/traces stay outside [Verified README]
- identity-shaping-framework repo Apache-2.0 last activity 2026-07-16 · 7 GitHub stars on 2026-09-24
Toolkit to post-train a model toward a described identity via data synthesis, evals and training, driven by a coding agent [Verified README]
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
5 of 5 lenses: evals, local-models, verification, agents
model-mirror (MIT) is adoptable now for pinning and hash-verifying local MLX/HF weights; speechmap-eval's judge-training and judge-evaluation tooling is a working example of validating an LLM judge; aibb is a clean pattern for publishing multi-model agent transcripts as a git record.
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
- Pilot model-mirror for pinning local model weights (github.com/xlr8harder/model-mirror). Commit-pinned, hash-verified mirrors keep local weights from updating silently Effort: small.
- Read speechmap-eval judge training and judge evaluation (github.com/xlr8harder/speechmap-eval). Concrete, open pipeline for calibrating an LLM judge at scale (845k responses) Effort: medium.
- Study aibb's harness/record split (github.com/xlr8harder/aibb). Public git record vs private traces/credentials is a pattern for publishing agent-fleet transcripts safely Effort: medium.
Before reusing anything: The listed repositories are MIT or Apache-2.0 [Verified: GitHub license API, 2026-09-25, evidence/licenses.jsonl]. SpeechMap response data contains third-party model outputs; provider terms may restrict reuse for training.
Sources
- speechmap.substack.com/p/debuting-speechmap-index-and-lab
- github.com/xlr8harder/speechmap-eval
- github.com/xlr8harder/speechmap-data
- github.com/xlr8harder/model-mirror
- github.com/xlr8harder/aibb
- github.com/xlr8harder/identity-shaping-framework
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 23 of people.jsonl.
Limits
Not verified: Did not run any of the tools. Did not read the SpeechMap judge methodology in detail, so its judge-accuracy claims are unverified. The SpeechMap site itself does not name the handle (rendered 2026-09-25), so it is cited only as the source of its own counts.
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.