Merge pull request #13 from Human-Augment-Analytics/rz/recruit1

Update some of the projects and recruitment statuses

Author: riyamzaman

projects/3d-fossils.yml

@@ -2,22 +2,24 @@ id: 3d-fossils
 name: 3D Fossils
 status: active      # active | completed | preprint | archived
 visibility: public # public | private
-recruiting: false
+recruiting: true
 faculty:
   - Arthur Porto
 researchers:
   - 
   - 
 advisors:
-  - 
-  - 
+  - Dr. Katherine Wolcott
 summary: >
-  Summary here.
+  The goal of this project is to leverage 3D generative AI for exploring vertebrae of both fossil and modern lizards. This project utilizes a Neural Shape Model (NSM), built on Meta AI’s DeepSDF (Deep Signed Distance Function). Our dataset consists of 3D surface models of over 2,000 vertebrae from a phylogenetically-diverse set of lizard species.
 tags:
   - 3D
+  - Machine Learning
+  - Image Processing
+  - Computer Graphics
 links:
   github: https://github.com/3D-fossils-Haag # placeholder, if public
-  docs: https://example.com/docs # Could be README, wiki, github pages, etc.
-  publication: https://doi.org/... # If exists/completed
-  forum: https://example.com/forum # link to forum - @James Hennessy
-  contact: mailto:lab@example.edu # Contact to reach out to for interested parties (lab-level email distribution)
+  docs: null # Could be README, wiki, github pages, etc.
+  publication: null # If exists/completed
+  forum: null # link to forum - @James Hennessy
+  contact: null # Contact to reach out to for interested parties (lab-level email distribution)

projects/3d-generative-models.yml

@@ -12,9 +12,11 @@ advisors:
   - 
   - 
 summary: >
-  Summary here.
+  This project aims to develop a state-of-the-art predictive tool that, given a partial or sparse 3D scan of a biological specimen, can generate a complete and anatomically coherent 3D structure. Moving beyond the linear limitations of traditional Statistical Shape Models (SSM) and PCA-based registration, this work will leverage the power of Conditional Denoising Diffusion Models (CDDMs).
+  The core of the project involves training a CDDM to learn the complex, non-linear deformation patterns required to warp a mean template shape to any complete sample in our dataset. By conditioning this process on a partial input scan, the model will learn to generate the most probable full deformation field, resulting in a high-fidelity shape completion. This generative approach is designed to produce reconstructions that are not only consistent with the provided data but are also more biologically plausible than those from methods constrained by linear assumptions.
 tags:
   - 3D
+  - Deep Generative Models
 links:
   github: https://github.com/alannadels/CDDM_Point_Set_Registration # placeholder, if public
   docs: https://example.com/docs # Could be README, wiki, github pages, etc.

projects/bird-audio.yml

@@ -2,7 +2,7 @@ id: bird-audio
 name: Bird Audio
 status: active      # active | completed | preprint | archived
 visibility: public # public | private
-recruiting: false
+recruiting: true
 faculty:
   - Benjamin Freeman
 researchers:

projects/bird-behavior.yml

@@ -2,7 +2,7 @@ id: bird-behavior
 name: Bird Behavior
 status: active      # active | completed | preprint | archived
 visibility: public # public | private
-recruiting: false
+recruiting: true
 faculty:
   - Benjamin Freeman
 researchers:

projects/inaturalist.yml

@@ -27,12 +27,12 @@ summary: >
   iNat's own platform, given community sensitivity around AI tools.
 tags:
     - iNaturalist
-    - citizen-science
-    - computer-vision
-    - machine-learning
-    - species-identification
-    - biodiversity
-    - annotator-reliability
+    - Citizen Science
+    - Computer Vision
+    - Machine Learning
+    - Species Identification
+    - Biodiversity
+    - Annotator Reliability
 links:
   github: https://github.gatech.edu/Mussmann-INaturalist
   docs: null

projects/knowledge-traceability.yml

@@ -9,10 +9,10 @@ researchers:
   - 
   - 
 advisors:
-  - 
-  - 
+  - Aaron Payne
+  - Raghu Mulukutla
 summary: >
-  Summary here.
+  The goal is to establish “Knowledge Traceability” (KT) as a novel methodology for scaling interdisciplinary research operations by publishing two synergistic papers. This initiative aims to transform project management from simple task tracking to “knowledge flow” monitoring, ensuring that daily research activities directly contribute to publishable outcomes.
 tags:
   - 
 links:

projects/lidar-stroud.yml

@@ -2,7 +2,7 @@ id: lidar-stroud
 name: LiDAR
 status: active      # active | completed | preprint | archived
 visibility: public # public | private
-recruiting: false
+recruiting: true
 faculty:
   - James Stroud
 researchers:
@@ -26,8 +26,8 @@ advisors:
 summary: >
   A tool that maps out Dr. Stroud's lizard island to aid in lizard evolution research. The tool should take in a point in space a return information about the vegetation in the area. The island is mapped using LiDAR scan data.
 tags:
-  - lidar
-  - qsm
+  - LiDAR
+  - QSM
 links:
   github: https://github.com/Landscape-CV/Ecomodel
   docs: null

projects/lizard-toe-pad.yml

@@ -2,7 +2,7 @@ id: lizard-toe-pad
 name: Lizard Toe Pad
 status: active      # active | completed | preprint | archived
 visibility: public # public | private
-recruiting: false
+recruiting: true
 faculty:
   - James Stroud
 researchers:

projects/llm-interpretability.yml

@@ -12,9 +12,14 @@ advisors:
   - 
   - 
 summary: >
-  Summary here.
+  Large language models (LLMs) have achieved remarkable performance across diverse tasks, yet their opacity presents significant challenges for deployment in high-stakes domains such as medicine and law, where explainability is essential. Traditional interpretability methods that examine model internals—including attention mechanisms and gradient analyses—are unavailable for closed APIs and often inadequately capture the complex, emergent behaviors characteristic of large-scale models. Currently, we lack robust tools to predict when or why an LLM will exhibit specific behaviors.
+  This project addresses these limitations through a comprehensive model-agnostic interpretability framework that operates without access to internal architecture or weights.
 tags:
-  - 
+  - Machine Learning
+  - Deep Learning
+  - AI
+  - Statistical Analysis
+  - Hypothesis Testing
 links:
   github: https://github.com/Human-Augment-Analytics/llm-fine-tuning-bias # placeholder, if public
   docs: https://example.com/docs # Could be README, wiki, github pages, etc.

projects/nahpu.yml

@@ -2,7 +2,7 @@ id: nahpu
 name: NAtural History Project Utility (NAHPU)
 status: active      # active | completed | preprint | archived
 visibility: public # public | private
-recruiting: false
+recruiting: true
 faculty:
   - Heru Handika
 researchers: