Humanity's Last Hackathon

community

Activity Feed Request to join this org

AI & ML interests

None defined yet.

Recent Activity

sum-guy-429 new activity 1 day ago

humanitys-last-hackathon/signup:docs: document Space Secrets / env vars for operators

jagdish-007 new activity 3 months ago

humanitys-last-hackathon/signup:FAQ: Is this REALLY mankinds last hackathon?

omarirfa new activity 4 months ago

humanitys-last-hackathon/README:is it time yet

View all activity

sum-guy-429

in humanitys-last-hackathon/signup 1 day ago

docs: document Space Secrets / env vars for operators

#5 opened 1 day ago by

sum-guy-429

epappas

authored 2 papers 8 days ago

Privacy Failure in Split-LLM Training, The Returned Gradient Nullifies the Decoys

Paper • 2609.04382 • Published 14 days ago • 5

Covenant-72B: Pre-Training a 72B LLM with Trustless Peers Over-the-Internet

Paper • 2603.08163 • Published Mar 9 • 5

codelion

authored 4 papers about 2 months ago

FAQ: Is this REALLY mankinds last hackathon?

#3 opened 5 months ago by

TheSystemGuy

codelion

posted an update 3 months ago

Post

645

SPROG-9M — a 9.37M parameter model trained from scratch to solve GSM8K-style math without using an LLM at inference.

The model, codelion/sprog-9m, predicts symbolic programs over number slots, then a deterministic executor does the arithmetic. With a simple verifier, it reaches ~11.8% on GSM8K test.

We also released the dataset: codelion/gsm8k-synth, 117K validated synthetic GSM8K-style problems.

Tiny model, no pretraining, no LLM at inference, runs on a laptop.

omarirfa

in humanitys-last-hackathon/README 4 months ago

is it time yet

🤗 8

#1 opened 4 months ago by

juiceb0xc0de

codelion

posted an update 4 months ago

Post

3271

Inspired by the Nemotron Diffusion recipe, check out dhara-250m: a 250M experimental language model that supports three decoding modes from one set of weights: autoregressive, block-diffusion, and self-speculation.

It is small, easy to try, and meant for exploring diffusion-style decoding and latency tradeoffs in compact LMs.

Model: codelion/dhara-250m

Try the chat demo here: codelion/dhara-chat