rbroc echo issues - Githubissues

rbroc / echo

A Scalable and Explainable Approach to Discriminating Between Human and Artificially Generated Text

https://cc.au.dk/en/clai/current-projects/a-scalable-and-explainable-approach-to-discriminating-between-human-and-artificially-generated-text

2 stars 1 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

Better table: Prelim results still, but a better overview of the results

#82 MinaAlmasi closed 1 week ago
0
Llm detector: add script for both zero-shot and few-shot learning

#81 MinaAlmasi closed 1 week ago
0
Baselines: add results for embeddings

#80 MinaAlmasi closed 2 weeks ago
2
Project Overview

#79 rbroc opened 3 weeks ago
0
feat: add feature importances script (on XGBOOST importances)

#78 MinaAlmasi closed 1 month ago
0
Start draft on Overleaf

#77 rbroc opened 1 month ago
0
Human baseline

#76 rbroc opened 1 month ago
1
Baselines

#75 rbroc opened 1 month ago
3
Features: Finalize feature sets

#74 rbroc opened 1 month ago
6
major refactoring: process and split data properly, save splitted data

#73 MinaAlmasi closed 1 month ago
0
Refactoring: Delete PROMPT_SELECT src and its results + some unused scripts in analysis

#72 MinaAlmasi closed 2 months ago
0
Refactoring: Delete some folders

#71 MinaAlmasi closed 1 month ago
2
Create script to split data for classifiers and save to folder

#70 MinaAlmasi closed 1 month ago
1
minor clean: delete rogue file

#69 MinaAlmasi closed 2 months ago
0
Refactoring: Small fixes to utils to ensure old code works + newer (prettier) tables

#68 MinaAlmasi closed 2 months ago
0
small fixes, docs to classify

#67 MinaAlmasi closed 2 months ago
0
Classify pipeline

#66 MinaAlmasi closed 2 months ago
0
Dropping raw features with many zero values prior to PCA?

#65 MinaAlmasi closed 1 month ago
1
Classify: Fix filtering lengths and identification of NA features

#64 MinaAlmasi closed 3 months ago
0
fix classify pipeline: prelim results on all PC comps

#63 MinaAlmasi closed 3 months ago
0
Computing Perplexity outside of TextDescriptives (and Entropy)

#62 MinaAlmasi closed 1 month ago
2
Refactoring: Weird steps in the pipeline (cleaning at various steps that could be streamlined)

#61 MinaAlmasi closed 1 month ago
1
Preliminary Classification Pipeline

#59 MinaAlmasi closed 7 months ago
0
DailyDialog: regenerate + extract new metrics

#58 MinaAlmasi closed 7 months ago
0
METRICS EXTRACTION + PERFORMANCE: Created one pipeline for standardising the extraction of both AI and human metrics + performance update.

#57 MinaAlmasi closed 7 months ago
0
DailyDialog: Regenerate dataset with correct lengths + extract new metrics for it

#56 MinaAlmasi closed 7 months ago
0
spring cleaning: adding documentation, removing old code, trim down folders

#55 MinaAlmasi closed 8 months ago
0
Preliminary analysis (checking lengths, PCA on simple features)

#54 MinaAlmasi closed 8 months ago
0
re Features (#74): features quantifying overlap / relation between source and completion

#53 rbroc closed 3 weeks ago
1
flag weird model-generated text

#52 rbroc closed 5 months ago
0
add stories 22 data for llama7b

#50 MinaAlmasi closed 8 months ago
0
Finish generations of data (for now!)

#49 MinaAlmasi closed 8 months ago
0
Final Prompts + Generation of Data

#48 MinaAlmasi closed 8 months ago
0
Landing on prompts

#47 MinaAlmasi closed 8 months ago
1
Clean human data for better model prompts

#46 MinaAlmasi closed 9 months ago
0
Human Data Cleaning: reformatting speakers in DailyDialog + inspect other human data

#45 MinaAlmasi closed 9 months ago
0
re Datasets (#2) - dailymail_cnn: weird cleaning or weird formatting?

#44 MinaAlmasi closed 3 weeks ago
8
Exploring Data BonusInfo in

#43 MinaAlmasi closed 9 months ago
0
Generating data with vLLM (generations + hacky fix to min tokens)

#41 MinaAlmasi closed 9 months ago
0
Minor update: Add temperature as CLI arg + add all sample params as dict in df col for generations

#40 MinaAlmasi closed 9 months ago
0
Dailydialog: Re-introduce [EOT] tokens as alternating speaker 1 and speaker 2 (+ general streamlining of data cleaning)

#39 MinaAlmasi closed 9 months ago
0
vLLM implementation

#38 MinaAlmasi closed 9 months ago
0
Fix Quantized Mdl Implementation, add probabilistic decoding, streamline generate scripts

#37 MinaAlmasi closed 10 months ago
0
Docs: Add more content to readme

#36 MinaAlmasi closed 11 months ago
0
MORE refactoring

#35 MinaAlmasi closed 11 months ago
0
(MINOR) restructuring/cleaning of repo

#34 MinaAlmasi closed 12 months ago
0
INTERACTIVE PLOTS + MODIFIED GEN PIPELINE

#33 MinaAlmasi closed 1 year ago
0
data cleanup before fitting models

#32 rbroc closed 1 month ago
2
PROMPT SELECTION: RESULTS README

#31 MinaAlmasi closed 1 year ago
0
PROMPT SELECTION: pca, euclidean distances

#30 MinaAlmasi closed 1 year ago
0