Mathieu Acher
@macher.bsky.social
Chess-loving professor and researcher who champion the integration of software engineering and AI for reproducible science.
Diving deep into software variability spaces, from Airbus to Linux.
@rennesuniv.bsky.social #INSA #IUF @InstUnivFr @Inria #IRISA
Diving deep into software variability spaces, from Airbus to Linux.
@rennesuniv.bsky.social #INSA #IUF @InstUnivFr @Inria #IRISA
I presented "Teaching Reproducibility and Embracing Variability: From Floating-Point Experiments to Replicating Research" at ACM REP conference 2025 .
Blog post with links to preprint, slides, and raw transcript: blog.mathieuacher.com/TeachingRepr...
Blog post with links to preprint, slides, and raw transcript: blog.mathieuacher.com/TeachingRepr...
July 31, 2025 at 9:30 AM
I presented "Teaching Reproducibility and Embracing Variability: From Floating-Point Experiments to Replicating Research" at ACM REP conference 2025 .
Blog post with links to preprint, slides, and raw transcript: blog.mathieuacher.com/TeachingRepr...
Blog post with links to preprint, slides, and raw transcript: blog.mathieuacher.com/TeachingRepr...
The latest generation of reasoning LLMs perform worse at #Chess compared to previous models. o3 & o4‑mini vs weak Stockfish: illegal moves in 88% & 94% of 67 games. o3 breaks rules in 4 moves; both resigned while winning. Worse than GPT‑3.5‑turbo‑instruct (1750 Elo)
June 26, 2025 at 3:31 PM
The latest generation of reasoning LLMs perform worse at #Chess compared to previous models. o3 & o4‑mini vs weak Stockfish: illegal moves in 88% & 94% of 67 games. o3 breaks rules in 4 moves; both resigned while winning. Worse than GPT‑3.5‑turbo‑instruct (1750 Elo)
Un élément nouveau de la vidéo #Devoxx concerne ce comportement étrange de gpt-3.5-turbo-instruct. A voir s'il est possible de reproduire ;) Assez lié à une autre série d'expériences où j'ai montré comment gagner en 4 ou 7 coups de manière systématique blog.mathieuacher.com/ChessWinning... 3/3
May 10, 2025 at 9:34 PM
Un élément nouveau de la vidéo #Devoxx concerne ce comportement étrange de gpt-3.5-turbo-instruct. A voir s'il est possible de reproduire ;) Assez lié à une autre série d'expériences où j'ai montré comment gagner en 4 ou 7 coups de manière systématique blog.mathieuacher.com/ChessWinning... 3/3
Real position coming from an online real game in #Chess960 I just played. Is it a draw? -0.3 according to Stockfish, but no clear plan. Chess engines are notoriously bad at resolving/assessing fortress-like position. But is it such a case? What do you think? #ChessEveryWhere
April 30, 2025 at 12:18 PM
Real position coming from an online real game in #Chess960 I just played. Is it a draw? -0.3 according to Stockfish, but no clear plan. Chess engines are notoriously bad at resolving/assessing fortress-like position. But is it such a case? What do you think? #ChessEveryWhere
🔎 A Chess Mystery
These mirrored positions should have the same evaluation, but at depth=20:
📊 Left: +0.66
📊 Right: -2.17
This is not just a low-depth issue—it rings a bell.
These mirrored positions should have the same evaluation, but at depth=20:
📊 Left: +0.66
📊 Right: -2.17
This is not just a low-depth issue—it rings a bell.
March 20, 2025 at 10:41 AM
🔎 A Chess Mystery
These mirrored positions should have the same evaluation, but at depth=20:
📊 Left: +0.66
📊 Right: -2.17
This is not just a low-depth issue—it rings a bell.
These mirrored positions should have the same evaluation, but at depth=20:
📊 Left: +0.66
📊 Right: -2.17
This is not just a low-depth issue—it rings a bell.
Metamorphic Testing, Reproducibility & a Curious Chess Engine Mystery
We replicated a study that found inconsistencies in how Stockfish (the best engine in the world) evaluates mirrored positions. But the key issue? Depth sensitivity.🧵⬇️
We replicated a study that found inconsistencies in how Stockfish (the best engine in the world) evaluates mirrored positions. But the key issue? Depth sensitivity.🧵⬇️
March 20, 2025 at 10:41 AM
Metamorphic Testing, Reproducibility & a Curious Chess Engine Mystery
We replicated a study that found inconsistencies in how Stockfish (the best engine in the world) evaluates mirrored positions. But the key issue? Depth sensitivity.🧵⬇️
We replicated a study that found inconsistencies in how Stockfish (the best engine in the world) evaluates mirrored positions. But the key issue? Depth sensitivity.🧵⬇️
Size coding an impressive graphics demo with wave simulations, sky and water shaders, etc. in 256 bytes. 256 b-y-t-e-s. Using a Rust-like syntax, some clever algos, and Web assembly. It's not my creation, and I found it in a video by
@lauriewired.bsky.social. Some notes below
@lauriewired.bsky.social. Some notes below
March 18, 2025 at 7:23 PM
Size coding an impressive graphics demo with wave simulations, sky and water shaders, etc. in 256 bytes. 256 b-y-t-e-s. Using a Rust-like syntax, some clever algos, and Web assembly. It's not my creation, and I found it in a video by
@lauriewired.bsky.social. Some notes below
@lauriewired.bsky.social. Some notes below
On a notamment trouvé un Jean-Jacques spécial... une variante très rare.
Article ici : inria.hal.science/hal-01104797
Article ici : inria.hal.science/hal-01104797
February 19, 2025 at 6:59 PM
On a notamment trouvé un Jean-Jacques spécial... une variante très rare.
Article ici : inria.hal.science/hal-01104797
Article ici : inria.hal.science/hal-01104797
It is the longest game out of 58 against a modest Stockfish. The opening is OKish, but then 6 pieces are given in a row! Final move: an illegal one. 4/6
January 27, 2025 at 4:42 PM
It is the longest game out of 58 against a modest Stockfish. The opening is OKish, but then 6 pieces are given in a row! Final move: an illegal one. 4/6