issues
search
danton-nlp
/
GEF
1
stars
0
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
Computed various extrinsic entity metrics #116
#133
anton164
closed
2 years ago
1
Update Section 4 (Evaluation)
#132
anton164
closed
2 years ago
0
Updated Figure 1
#131
anton164
closed
2 years ago
0
If time: do stats
#130
anton164
closed
2 years ago
0
Familiarize ourselves with EMNLP Submission Requirements
#129
anton164
closed
2 years ago
0
Clean up repository and README
#128
anton164
closed
2 years ago
0
Anonymize the paper
#127
anton164
closed
2 years ago
0
Match entities by word overlap instead of string-contains ("Ola" in "Tshibola" case)
#126
anton164
opened
2 years ago
0
Added full meng results & annotated first 25 with strict intrinsic
#125
anton164
closed
2 years ago
0
Iterations Analysis
#124
dleve123
closed
2 years ago
1
Audit & Fix bad annotations
#123
dleve123
closed
2 years ago
0
Updated logs for full dataset (bart & pegasus)
#122
anton164
closed
2 years ago
1
Added fallback for failed summary generation and updated results #113
#121
anton164
closed
2 years ago
0
Simplified & sped up evaluation script by moving rouge score computation to separate script
#120
anton164
closed
2 years ago
0
Added intrinsic option to annotation UI & fixed missing Unknown option
#119
anton164
closed
2 years ago
0
Add intrinsic option in annotation UI to force intrinsic despite entity not being in source
#118
anton164
closed
2 years ago
2
Simplified annotation UI to simply select factual/non-factual fixes #61
#117
anton164
closed
2 years ago
0
Add Number of Extrinsic Entities to Main Results
#116
anton164
closed
2 years ago
0
Daniel annotates Pegasus to 100 (baseline / oracle & classifier)
#115
anton164
closed
2 years ago
1
Anton annotates King
#114
anton164
closed
2 years ago
0
Investigate how we fall back when summary generation fails
#113
anton164
closed
2 years ago
1
Fix FBS: Jean-Claude Duncker and Sir Bernard Hogan-Howe example
#112
anton164
closed
2 years ago
0
Fix iterative_constraints: consider a summary completed if there are no new constraints
#111
anton164
closed
2 years ago
1
Added plot for word edit distance (in iteration_changes notebook)
#110
anton164
closed
2 years ago
0
Compute Rouge Scores
#109
anton164
closed
2 years ago
1
Compute extrinsic factuality ratio
#108
dleve123
closed
2 years ago
2
Data Bug: Ensure that all in_source annotations are either Non-Hallucinated or Intrinsic Hallucination
#107
dleve123
closed
2 years ago
2
Run FBS and Pegasus on the full dataset
#106
anton164
closed
2 years ago
2
Finalize Analysis - What changes does FBS make?
#105
anton164
closed
2 years ago
1
v2 draft of intro
#104
anton164
closed
2 years ago
0
Entity Type Analysis Plot & Writing
#103
anton164
closed
2 years ago
0
Extrinsic Hallucinations Ratio for Each System
#102
anton164
closed
2 years ago
5
Intro & Related Work [Full Draft]
#101
anton164
closed
2 years ago
2
Draft EMNLP Submission of Abtract
#100
anton164
closed
2 years ago
1
Support for varying beam size
#99
anton164
closed
2 years ago
1
More annotations
#98
dleve123
closed
2 years ago
0
Updated iteration change logs (without skips)
#97
anton164
closed
2 years ago
0
Default to not counting skips in evaluation (added tests)
#96
anton164
closed
2 years ago
1
Added Meng et RL-generated summaries
#95
anton164
closed
2 years ago
0
Logs for showing how FBS summaries change over each iteration
#94
anton164
closed
2 years ago
0
v2 draft of main results section
#93
anton164
closed
2 years ago
0
v1 draft of analysis section
#92
anton164
closed
2 years ago
0
Update FBS -> GEF
#91
anton164
closed
2 years ago
0
Send paper draft to Pagnoni
#90
anton164
closed
2 years ago
0
Merge Figure 1 & Table 1
#89
anton164
closed
2 years ago
0
Merge evaluation table 3 & 4
#88
anton164
closed
2 years ago
1
Add PEGASUS to evaluation table
#87
anton164
closed
2 years ago
1
Model comparisons: clearly argue why we don't compare against FROST
#86
anton164
closed
2 years ago
1
Method section: shorten beam search section and elevate it to a higher level instead of describing beam search details
#85
anton164
closed
2 years ago
0
Method section: Simplify notation
#84
anton164
closed
2 years ago
0
Previous
Next