Brand v2: banner, eyebrow, About footer, 'Where it falls short'; Udgam Labs name, lowercase ids

README.md

@@ -5,6 +5,7 @@ language:
 pipeline_tag: text-generation
 library_name: transformers
 tags:
+- udgamlabs
 - math
 - word-problems
 - program-aided
@@ -19,6 +20,10 @@ datasets:
 - openai/gsm8k
 ---
 
+![Lilavati · Udgam Labs](./banner.png)
+
+**UDGAM LABS** · Maths model · 95M parameters
+
 # Lilavati: Pocket Mathematician
 
 **A 95M-parameter English model, trained from scratch by Udgam Labs, that solves maths word problems by writing a short
@@ -27,7 +32,7 @@ Python program and running it.** It is small enough to run on a laptop CPU.
 The name comes from *Līlāvatī* (लीलावती), the book of arithmetic word problems that the mathematician Bhāskara II wrote
 in 1150 CE.
 
-**More:** [try it in your browser](https://udgamlabs.com/models/lilavati) · [code and evaluation on GitHub](https://github.com/UdgamLabs/lilavati) · [the checker model](https://huggingface.co/UdgamLabs/Lilavati-Checker)
+**More:** [try it in your browser](https://udgamlabs.com/models/lilavati) · [code and evaluation on GitHub](https://github.com/udgamlabs/lilavati) · [the checker model](https://huggingface.co/udgamlabs/Lilavati-Checker)
 
 ```
 Question: A chai stall sells 45 cups of tea in the morning and twice as many in the evening.
@@ -49,7 +54,7 @@ Lilavati writes:                                         Running it gives:
 | Context | 4,096 tokens |
 | Input / output | an English word problem → a Python function whose return value is the answer |
 | Training | from scratch: public synthetic maths datasets with Python solutions, plus general English text; then further training on word problems with execution-verified Python solutions |
-| Companion | [Lilavati-Checker](https://huggingface.co/UdgamLabs/Lilavati-Checker) (95M) rates programs so that many tries can be combined |
+| Companion | [Lilavati-Checker](https://huggingface.co/udgamlabs/Lilavati-Checker) (95M) rates programs so that many tries can be combined |
 | Licence | Apache-2.0 |
 
 ## Quick start
@@ -58,12 +63,12 @@ The `lilavati` package writes the programs, runs them in a restricted sandbox an
 
 ```bash
 pip install torch transformers
-git clone https://github.com/UdgamLabs/lilavati && cd lilavati && pip install -e .
+git clone https://github.com/udgamlabs/lilavati && cd lilavati && pip install -e .
 ```
 
 ```python
 from lilavati import Lilavati
-lv = Lilavati()      # downloads UdgamLabs/Lilavati and UdgamLabs/Lilavati-Checker
+lv = Lilavati()      # downloads udgamlabs/Lilavati and udgamlabs/Lilavati-Checker
 
 r = lv.solve("A shop sells pencils at 8 rupees each. Riya buys 6 pencils and 2 erasers at 5 rupees each. "
              "She pays with a 100 rupee note. How much change does she get?")
@@ -170,6 +175,19 @@ overlap against the full sets: GSM-Plus 39.1 vs 39.2, SVAMP 69.6 vs 69.9 and ASD
 does not explain the results. A 13-gram check finds near-copies, not reworded problems. Problems added in later training
 were filtered against all these test sets (8-word sequences).
 
+## Where it falls short
+
+All numbers are from the tables above.
+
+- **Perturbed and harder problems.** One greedy try scores 54.7 on GSM8K test but 39.2 on GSM-Plus (the 7 answerable
+  perturbation types) and 43.7 on GSM-Hard.
+- **Against TinyGSM.** TinyGSM's 125M model, first pretrained on Phi-1.5 data, reports 63.1 with one try (Lilavati 54.7)
+  and 68.9 when its verifier picks one of 48 programs (Lilavati with its checker, same protocol: 66.7). Those weights are
+  not released.
+- **Choosing among tries.** At least one of 48 programs is right on 89.5% of GSM8K problems, but the checker-weighted vote
+  finds the right answer on 72.3%.
+- The kinds of problem it gets wrong are listed under *What it is good at, and what it is not*.
+
 ## How it works
 
 <p align="center"><img src="assets/tries.png" alt="Eight programs for one cricket question; five return 75, and the checker-weighted vote picks 75" width="100%"></p>
@@ -212,6 +230,10 @@ Weights and code: Apache-2.0.
   title  = {Lilavati: a 95M pocket mathematician that solves word problems by writing Python},
   author = {Udgam Labs},
   year   = {2026},
-  url    = {https://huggingface.co/UdgamLabs/Lilavati}
+  url    = {https://huggingface.co/udgamlabs/Lilavati}
 }
 ```
+
+---
+
+**About Udgam Labs.** Udgam Labs is an independent AI research lab. We work from first principles: before we build, we understand why things work, whether we are training a model, fine-tuning one or applying AI to a new problem. Udgam (उद्गम) is Sanskrit for origin. [udgamlabs.com](https://udgamlabs.com)