Spaces:

cdpearlman
/

LLMVis

Sleeping

App Files Files Community

LLMVis

Commit History

Apr 26

03:04
cleanup

cdpearlman

1b3b7d0
03:03
Tutorial, feedback, and github buttons added

cdpearlman

e1f91c2

Apr 22

18:24
Fixed slider loading bug

cdpearlman

4416500

Mar 31

20:16
Updated and partitioned glossary

cdpearlman

e2d10e5
15:16
Moved glossary button to header instead of hidden in sidebar

cdpearlman

42ae447
15:14
fixed ablation bug, ablated output was shown in pipeline

cdpearlman

96418df

Mar 24

17:13
Updated ContextKit skills and memory system

cdpearlman

a836411

Mar 22

00:46
Fixed ablation experiments, drilled down on which pre-loaded ones to run

cdpearlman

3d71d04

Mar 21

23:20
Added simpler ablation experiment buttons to add full categories

cdpearlman

3c5f2fd
21:33
Added pre-built prompts for app and chatbot

cdpearlman

018a9f2

Mar 19

22:53
memory update

cdpearlman

bff3e4d
22:42
Bug fix: type issue with models leading to gibberish outputs

cdpearlman

5528a77
19:28
Replace token references with word in output selection stage⁠

cdpearlman +1: Claude Opus 4.6

ac0335e
19:01
Jargon reduction memory update

cdpearlman

7b0e148

Mar 17

15:24
Rename L#-H# notation to L#-D# for detector consistency⁠

cdpearlman +1: Claude Opus 4.6

deb2071

Mar 16

17:23
Reduce jargon across UI for younger/less-technical audiences⁠

cdpearlman +1: Claude Opus 4.6

abf6a1c
17:05
Improved rules and added claude.md

cdpearlman

2f024ca

Mar 6

20:59
Fixed chatbot bubble generation bug

cdpearlman

d075209
20:45
Added numbers to BertViz visualization, easier to map attention head numbers

cdpearlman

67f5be3
20:34
Fixed single-token sequence UI bug

cdpearlman

e248349
19:50
Fixed ablation original prompt generation bug

cdpearlman

1356785
19:37
Fixed ablation graphs first token population, still need to fix original graphs when no beam selected

cdpearlman

cb213d9
19:13
Fixed ablation graphs first token population, still need to fix original graphs when no beam selected

cdpearlman

670d882

Mar 5

01:57
Fixed duplicate activation store bug

cdpearlman

78cc5b4
01:50
added color swatches for easy mapping from attention categories to BertViz

cdpearlman

96e2bc7
01:30
Fixed model-switching bug with activation store clearing

cdpearlman

7ad3e7b

Mar 2

20:42
ContextKit LLM memory update

cdpearlman

11aaea3

Feb 27

20:48
display attention category scores in order

cdpearlman

c629c1f
19:59
Added loading screen to pipeline steps for better UX

cdpearlman

93c55d0
19:09
completed the rest of the explanation improvements from JARVIS audit

cdpearlman

ff1ba74
01:38
New models chosen and rag docs updated

cdpearlman

f478beb

Feb 26

21:08
Attention refactor, better categorization and explanation

cdpearlman

7fa8fb4
19:30
plans update

cdpearlman

d86e476

Feb 25

02:18
Fixed glossary UI to be cleaner

cdpearlman

1cce926
01:56
Added glossary embeddings and explanations

cdpearlman

2800225

Feb 22

02:49
Chatbot improvements: UI cleaning, chat streaming, typing animation, and better markdown formatting

cdpearlman

a684a8d
02:21
ablation experiment fixes, better UI and full sequence comparison

cdpearlman

a938eeb

Feb 16

21:10
feat(output): Add token scrubber with per-position top-5 next-token probabilities⁠

cdpearlman +1: Cursor

d60cfe2
20:49
Added full sequence to analysis, still need to do output scrubber

cdpearlman

3f991b4
20:27
Add drag-to-resize on chat window left edge⁠

cdpearlman +1: Cursor

718dfcb

Feb 5

21:03
Add 30 RAG documents for chatbot knowledge base⁠

cdpearlman +1: Cursor

78691d1
20:13
Add Hugging Face Spaces configuration to README⁠

cdpearlman +1: Cursor

e582564
20:06
Add Dockerfile for Hugging Face Spaces deployment

cdpearlman

159696b

Feb 4

22:27
Added message greeting to chatbot

cdpearlman

b6266bb
22:08
Reset chat history on browser refresh to show greeting⁠

cdpearlman +1: Cursor

cbe8dbf
22:07
Add greeting message to chatbot on first open⁠

cdpearlman +1: Cursor

678a73d

Feb 3

21:22
cleanup stuff

cdpearlman

1f487a9
20:56
Switch to cost-optimized paid models: Gemini 2.5 Flash Lite + text-embedding-3-small⁠

cdpearlman +1: Cursor

129c11e
20:37
Switch to free models: Qwen3-Next chat + local embeddings⁠

cdpearlman +1: Cursor

94b077a
20:20
Migrate from Gemini to OpenRouter API⁠

cdpearlman +1: Cursor

ef3e36a

Commit History

Apr 26

cleanup

Tutorial, feedback, and github buttons added

Apr 22

Fixed slider loading bug

Mar 31

Updated and partitioned glossary

Moved glossary button to header instead of hidden in sidebar

fixed ablation bug, ablated output was shown in pipeline

Mar 24

Updated ContextKit skills and memory system

Mar 22

Fixed ablation experiments, drilled down on which pre-loaded ones to run

Mar 21

Added simpler ablation experiment buttons to add full categories

Added pre-built prompts for app and chatbot

Mar 19

memory update

Bug fix: type issue with models leading to gibberish outputs

Replace token references with word in output selection stage⁠

Jargon reduction memory update

Mar 17

Rename L#-H# notation to L#-D# for detector consistency⁠

Mar 16

Reduce jargon across UI for younger/less-technical audiences⁠

Improved rules and added claude.md

Mar 6

Fixed chatbot bubble generation bug

Added numbers to BertViz visualization, easier to map attention head numbers

Fixed single-token sequence UI bug

Fixed ablation original prompt generation bug

Fixed ablation graphs first token population, still need to fix original graphs when no beam selected

Fixed ablation graphs first token population, still need to fix original graphs when no beam selected

Mar 5

Fixed duplicate activation store bug

added color swatches for easy mapping from attention categories to BertViz

Fixed model-switching bug with activation store clearing

Mar 2

ContextKit LLM memory update

Feb 27

display attention category scores in order

Added loading screen to pipeline steps for better UX

completed the rest of the explanation improvements from JARVIS audit

New models chosen and rag docs updated

Feb 26

Attention refactor, better categorization and explanation

plans update

Feb 25

Fixed glossary UI to be cleaner

Added glossary embeddings and explanations

Feb 22

Chatbot improvements: UI cleaning, chat streaming, typing animation, and better markdown formatting

ablation experiment fixes, better UI and full sequence comparison

Feb 16

feat(output): Add token scrubber with per-position top-5 next-token probabilities⁠

Added full sequence to analysis, still need to do output scrubber

Add drag-to-resize on chat window left edge⁠

Feb 5

Add 30 RAG documents for chatbot knowledge base⁠

Add Hugging Face Spaces configuration to README⁠

Add Dockerfile for Hugging Face Spaces deployment

Feb 4

Added message greeting to chatbot

Reset chat history on browser refresh to show greeting⁠

Add greeting message to chatbot on first open⁠

Feb 3

cleanup stuff

Switch to cost-optimized paid models: Gemini 2.5 Flash Lite + text-embedding-3-small⁠

Switch to free models: Qwen3-Next chat + local embeddings⁠

Migrate from Gemini to OpenRouter API⁠