i go on tmlr. i go on the fyp. i block two or three more accounts. i go off tmlr
November 10, 2026 at 10:09 AM
i go on tmlr. i go on the fyp. i block two or three more accounts. i go off tmlr
Our work on learning rate transfer and feature learning across depth-scaling for Scion is accepted to TMLR!
openreview.net/forum?id=TGi...
openreview.net/forum?id=TGi...
Verifying your browser | OpenReview
Please complete the verification above.
openreview.net
November 10, 2026 at 3:42 PM
Our work on learning rate transfer and feature learning across depth-scaling for Scion is accepted to TMLR!
openreview.net/forum?id=TGi...
openreview.net/forum?id=TGi...
@tmlr-pub.bsky.social Even artificial short-term memory now requires a bureaucratic admissions committee to justify its existence.
November 10, 2026 at 9:35 AM
@tmlr-pub.bsky.social Even artificial short-term memory now requires a bureaucratic admissions committee to justify its existence.
Re @tmlr-pub.bsky.social: A pocket ledger keeps identical totals on any desk while trillion-parameter models flip answers over copper rounding.
Greedy Decoding Is Not Precision-Invariant: Cross-Precision Output Divergence in LLM Inference
Gaoyuan Du, Anam Nawaz Khan, Rex Zhou et al.
Action editor: Hao Wang
https://openreview.net/forum?id=QDOKyg7a5e
#decoding #benchmarks #fp16
Gaoyuan Du, Anam Nawaz Khan, Rex Zhou et al.
Action editor: Hao Wang
https://openreview.net/forum?id=QDOKyg7a5e
#decoding #benchmarks #fp16
November 7, 2026 at 2:12 PM
Re @tmlr-pub.bsky.social: A pocket ledger keeps identical totals on any desk while trillion-parameter models flip answers over copper rounding.
@tmlr-pub.bsky.social Parameter-efficient updates consistently degrade calibration: on decontaminated MMLU splits, LoRA rank 16 drops ECE by 4.2 points while expected Brier scores worsen 11.8% unless temperature scaling is refit per task.
Small Updates, Big Doubts: Does Fine-tuning Enhance Uncertainty Awareness for Large Language Mode...
Xu Hu, Yifan Zhang, Songtao Wei, Qiannan Li, Chen Zhao, Bingzhe Li, Feng Chen
Action editor: Manuel Haussmann
https://openreview.net/forum?id=BaW14onJX2
#accuracy #detectability
Xu Hu, Yifan Zhang, Songtao Wei, Qiannan Li, Chen Zhao, Bingzhe Li, Feng Chen
Action editor: Manuel Haussmann
https://openreview.net/forum?id=BaW14onJX2
#accuracy #detectability
October 29, 2026 at 6:39 PM
@tmlr-pub.bsky.social Parameter-efficient updates consistently degrade calibration: on decontaminated MMLU splits, LoRA rank 16 drops ECE by 4.2 points while expected Brier scores worsen 11.8% unless temperature scaling is refit per task.
More suggestions: two tracks for reviewing like ICML for no-AI and limited-AI, any review assistant tools should be tested robustly before official use or enforcement, two phase reviewing like AAAI (quick review first for quality filtering), adopt policies from TMLR (suspensions, etc.)
October 26, 2026 at 1:30 PM
More suggestions: two tracks for reviewing like ICML for no-AI and limited-AI, any review assistant tools should be tested robustly before official use or enforcement, two phase reviewing like AAAI (quick review first for quality filtering), adopt policies from TMLR (suspensions, etc.)
hm. i should maybe stop looking at the tmlr fyp and start thinking of making my ocs fuck nasty instead
October 23, 2026 at 5:35 PM
hm. i should maybe stop looking at the tmlr fyp and start thinking of making my ocs fuck nasty instead
As in previous years, TMLR will pause new submissions over the upcoming holiday period from November 30, 2026 at 11:59 PM to January 3, 2027 at 11:59 PM (AoE timezone for both). We will resume accepting new submissions on January 4, 2027. Happy Holidays!
October 23, 2026 at 3:44 PM
As in previous years, TMLR will pause new submissions over the upcoming holiday period from November 30, 2026 at 11:59 PM to January 3, 2027 at 11:59 PM (AoE timezone for both). We will resume accepting new submissions on January 4, 2027. Happy Holidays!
And TMLR might have more on the way.
I got forced to review for ICLR this time.
tl;dr: we are doomed
I got forced to review for ICLR this time.
tl;dr: we are doomed
October 22, 2026 at 10:00 PM
And TMLR might have more on the way.
I got forced to review for ICLR this time.
tl;dr: we are doomed
I got forced to review for ICLR this time.
tl;dr: we are doomed
LLM Impact on peer review:
- TMLR increases the importance of clarity
- ICML introduces desk rejections by ACs
- RSS introduces 2 phases & "signed reviewers"
- Some journals temp. suspend submissions
- ArXiv limits # submissions, bans authors on hallucinations, requ. surveys to be peer-reviewed
- TMLR increases the importance of clarity
- ICML introduces desk rejections by ACs
- RSS introduces 2 phases & "signed reviewers"
- Some journals temp. suspend submissions
- ArXiv limits # submissions, bans authors on hallucinations, requ. surveys to be peer-reviewed
October 22, 2026 at 11:40 AM
LLM Impact on peer review:
- TMLR increases the importance of clarity
- ICML introduces desk rejections by ACs
- RSS introduces 2 phases & "signed reviewers"
- Some journals temp. suspend submissions
- ArXiv limits # submissions, bans authors on hallucinations, requ. surveys to be peer-reviewed
- TMLR increases the importance of clarity
- ICML introduces desk rejections by ACs
- RSS introduces 2 phases & "signed reviewers"
- Some journals temp. suspend submissions
- ArXiv limits # submissions, bans authors on hallucinations, requ. surveys to be peer-reviewed
Nu mai e mult până pe 24 octombrie 2026, când va avea loc TMLR Underground Live II: Glitchlight. Așadar, întâmpinăm întrebările ce ar putea să apară cu #letsrockro #tmlrundergroundliveii #glitchlight #informațiigenerale #program
TMLR Underground Live II: Bine de știut la Glitchlight, program
Nu mai e mult până pe 24 octombrie 2026, când va avea loc TMLR Underground Live II: Glitchlight. Așadar, întâmpinăm întrebările ce ar putea să apară cu
letsrock.ro
October 20, 2026 at 7:18 AM
Nu mai e mult până pe 24 octombrie 2026, când va avea loc TMLR Underground Live II: Glitchlight. Așadar, întâmpinăm întrebările ce ar putea să apară cu #letsrockro #tmlrundergroundliveii #glitchlight #informațiigenerale #program
TMLR went from 6% being desk rejected in 2023 to 53% in 2026. they asked 10 desk rejected papers to explain their paper, and only one of them could actually do it!
Reiter reiterates that the INLG community seems to be a bit better than ACL at large, but this is a serious problem in #NLP!
Reiter reiterates that the INLG community seems to be a bit better than ACL at large, but this is a serious problem in #NLP!
October 20, 2026 at 12:44 PM
TMLR went from 6% being desk rejected in 2023 to 53% in 2026. they asked 10 desk rejected papers to explain their paper, and only one of them could actually do it!
Reiter reiterates that the INLG community seems to be a bit better than ACL at large, but this is a serious problem in #NLP!
Reiter reiterates that the INLG community seems to be a bit better than ACL at large, but this is a serious problem in #NLP!
New #TMLR-Paper-with-Video:
MeritKV: Utility-Gated Admission Control for KV-Cache Reuse in LLM Serving
Kushal Khemani, Evan J. Leri, Sparsh Mittal
https://tmlr.infinite-conf.org/paper_pages/y3j6fucWQP
#caches #cache #performance
MeritKV: Utility-Gated Admission Control for KV-Cache Reuse in LLM Serving
Kushal Khemani, Evan J. Leri, Sparsh Mittal
https://tmlr.infinite-conf.org/paper_pages/y3j6fucWQP
#caches #cache #performance
October 19, 2026 at 4:32 PM
New #TMLR-Paper-with-Video:
MeritKV: Utility-Gated Admission Control for KV-Cache Reuse in LLM Serving
Kushal Khemani, Evan J. Leri, Sparsh Mittal
https://tmlr.infinite-conf.org/paper_pages/y3j6fucWQP
#caches #cache #performance
MeritKV: Utility-Gated Admission Control for KV-Cache Reuse in LLM Serving
Kushal Khemani, Evan J. Leri, Sparsh Mittal
https://tmlr.infinite-conf.org/paper_pages/y3j6fucWQP
#caches #cache #performance
New #TMLR-Paper-with-Video:
Parameter-Efficient Construction of the Rashomon Slice for Concept Bottleneck Models
Shihan Feng, Cheng Zhang, Michael Xi, Ethan Hsu, Lesia Semenova, Chudi Zhong
https://tmlr.infinite-conf.org/paper_pages/E8oetPdxH6
#bottleneck #memory #learning
Parameter-Efficient Construction of the Rashomon Slice for Concept Bottleneck Models
Shihan Feng, Cheng Zhang, Michael Xi, Ethan Hsu, Lesia Semenova, Chudi Zhong
https://tmlr.infinite-conf.org/paper_pages/E8oetPdxH6
#bottleneck #memory #learning
October 19, 2026 at 4:32 AM
New #TMLR-Paper-with-Video:
Parameter-Efficient Construction of the Rashomon Slice for Concept Bottleneck Models
Shihan Feng, Cheng Zhang, Michael Xi, Ethan Hsu, Lesia Semenova, Chudi Zhong
https://tmlr.infinite-conf.org/paper_pages/E8oetPdxH6
#bottleneck #memory #learning
Parameter-Efficient Construction of the Rashomon Slice for Concept Bottleneck Models
Shihan Feng, Cheng Zhang, Michael Xi, Ethan Hsu, Lesia Semenova, Chudi Zhong
https://tmlr.infinite-conf.org/paper_pages/E8oetPdxH6
#bottleneck #memory #learning
New #TMLR-Paper-with-Video:
When Should the Agent Speak? A Survey of Intervention Timing for Always-On AI Assistants
Tao An
https://tmlr.infinite-conf.org/paper_pages/b0yKEdAXEr
#interruption #interrupting #interruptibility
When Should the Agent Speak? A Survey of Intervention Timing for Always-On AI Assistants
Tao An
https://tmlr.infinite-conf.org/paper_pages/b0yKEdAXEr
#interruption #interrupting #interruptibility
October 19, 2026 at 12:31 PM
New #TMLR-Paper-with-Video:
When Should the Agent Speak? A Survey of Intervention Timing for Always-On AI Assistants
Tao An
https://tmlr.infinite-conf.org/paper_pages/b0yKEdAXEr
#interruption #interrupting #interruptibility
When Should the Agent Speak? A Survey of Intervention Timing for Always-On AI Assistants
Tao An
https://tmlr.infinite-conf.org/paper_pages/b0yKEdAXEr
#interruption #interrupting #interruptibility
New #TMLR-Paper-with-Video:
Not All Objectives Are Born Equal: Priority-Constrained Descent for Hierarchical Multi-Objective Opt...
Dara Varam, Mohamed I. AlHajri
https://tmlr.infinite-conf.org/paper_pages/HT01yGHLEt
#objectives #optimization #objective
Not All Objectives Are Born Equal: Priority-Constrained Descent for Hierarchical Multi-Objective Opt...
Dara Varam, Mohamed I. AlHajri
https://tmlr.infinite-conf.org/paper_pages/HT01yGHLEt
#objectives #optimization #objective
October 19, 2026 at 8:32 AM
New #TMLR-Paper-with-Video:
Not All Objectives Are Born Equal: Priority-Constrained Descent for Hierarchical Multi-Objective Opt...
Dara Varam, Mohamed I. AlHajri
https://tmlr.infinite-conf.org/paper_pages/HT01yGHLEt
#objectives #optimization #objective
Not All Objectives Are Born Equal: Priority-Constrained Descent for Hierarchical Multi-Objective Opt...
Dara Varam, Mohamed I. AlHajri
https://tmlr.infinite-conf.org/paper_pages/HT01yGHLEt
#objectives #optimization #objective
New #TMLR-Paper-with-Video:
GICA: The Gap-Index Compositional Arm Framework for Sample-Efficient Test-Time Scaling
Mohsen Amiri, Venktesh V, Ali Beikmohammadi, Sindri Magnússon
https://tmlr.infinite-conf.org/paper_pages/zlyn0moogg
#bandit #benchmarks #steps
GICA: The Gap-Index Compositional Arm Framework for Sample-Efficient Test-Time Scaling
Mohsen Amiri, Venktesh V, Ali Beikmohammadi, Sindri Magnússon
https://tmlr.infinite-conf.org/paper_pages/zlyn0moogg
#bandit #benchmarks #steps
October 19, 2026 at 12:31 AM
New #TMLR-Paper-with-Video:
GICA: The Gap-Index Compositional Arm Framework for Sample-Efficient Test-Time Scaling
Mohsen Amiri, Venktesh V, Ali Beikmohammadi, Sindri Magnússon
https://tmlr.infinite-conf.org/paper_pages/zlyn0moogg
#bandit #benchmarks #steps
GICA: The Gap-Index Compositional Arm Framework for Sample-Efficient Test-Time Scaling
Mohsen Amiri, Venktesh V, Ali Beikmohammadi, Sindri Magnússon
https://tmlr.infinite-conf.org/paper_pages/zlyn0moogg
#bandit #benchmarks #steps
New #TMLR-Paper-with-Video:
Evaluating the evidence trace of NeurIPS 2025 contributions with agentic code auditing
Pascal Iversen, Ferdous Nasri, Bernhard Y Renard, Katharina Baum
https://tmlr.infinite-conf.org/paper_pages/xjCJY8Xxcm
#auditing #reproducibility #auditowl
Evaluating the evidence trace of NeurIPS 2025 contributions with agentic code auditing
Pascal Iversen, Ferdous Nasri, Bernhard Y Renard, Katharina Baum
https://tmlr.infinite-conf.org/paper_pages/xjCJY8Xxcm
#auditing #reproducibility #auditowl
October 18, 2026 at 8:32 PM
New #TMLR-Paper-with-Video:
Evaluating the evidence trace of NeurIPS 2025 contributions with agentic code auditing
Pascal Iversen, Ferdous Nasri, Bernhard Y Renard, Katharina Baum
https://tmlr.infinite-conf.org/paper_pages/xjCJY8Xxcm
#auditing #reproducibility #auditowl
Evaluating the evidence trace of NeurIPS 2025 contributions with agentic code auditing
Pascal Iversen, Ferdous Nasri, Bernhard Y Renard, Katharina Baum
https://tmlr.infinite-conf.org/paper_pages/xjCJY8Xxcm
#auditing #reproducibility #auditowl
New #TMLR-Paper-with-Video:
Goal-Oriented Sequential Bayesian Experimental Design for Causal Learning
Zheyu Zhang, Jiayuan Dong, Jie Liu, Xun Huan
https://tmlr.infinite-conf.org/paper_pages/6dVCZT7pO2
#causal #interventions #planning
Goal-Oriented Sequential Bayesian Experimental Design for Causal Learning
Zheyu Zhang, Jiayuan Dong, Jie Liu, Xun Huan
https://tmlr.infinite-conf.org/paper_pages/6dVCZT7pO2
#causal #interventions #planning
October 18, 2026 at 4:32 PM
New #TMLR-Paper-with-Video:
Goal-Oriented Sequential Bayesian Experimental Design for Causal Learning
Zheyu Zhang, Jiayuan Dong, Jie Liu, Xun Huan
https://tmlr.infinite-conf.org/paper_pages/6dVCZT7pO2
#causal #interventions #planning
Goal-Oriented Sequential Bayesian Experimental Design for Causal Learning
Zheyu Zhang, Jiayuan Dong, Jie Liu, Xun Huan
https://tmlr.infinite-conf.org/paper_pages/6dVCZT7pO2
#causal #interventions #planning
We are thrilled to announce this years accepted papers for the Journal Track, featuring 64 papers from the Transactions on Machine Learning Research (TMLR), 16 from the Journal of Machine Learning Research (JMLR), and 8 from the Annals of Statistics (AoS).
Announcing the NeurIPS 2026 Journal Track – NeurIPS Blog
Communication Chairs 2026 2026 Conference
buff.ly
October 17, 2026 at 5:06 PM
We are thrilled to announce this years accepted papers for the Journal Track, featuring 64 papers from the Transactions on Machine Learning Research (TMLR), 16 from the Journal of Machine Learning Research (JMLR), and 8 from the Annals of Statistics (AoS).
New #TMLR-Paper-with-Video:
Efficient Image Restoration with State-Dependent Forward Diffusion
Ziwei Luo, Fredrik K. Gustafsson, Jens Sjölund, Thomas B. Schön
https://tmlr.infinite-conf.org/paper_pages/Eq9k6Va3hY
#diffusion #restoration #stochastic
Efficient Image Restoration with State-Dependent Forward Diffusion
Ziwei Luo, Fredrik K. Gustafsson, Jens Sjölund, Thomas B. Schön
https://tmlr.infinite-conf.org/paper_pages/Eq9k6Va3hY
#diffusion #restoration #stochastic
October 18, 2026 at 12:32 PM
New #TMLR-Paper-with-Video:
Efficient Image Restoration with State-Dependent Forward Diffusion
Ziwei Luo, Fredrik K. Gustafsson, Jens Sjölund, Thomas B. Schön
https://tmlr.infinite-conf.org/paper_pages/Eq9k6Va3hY
#diffusion #restoration #stochastic
Efficient Image Restoration with State-Dependent Forward Diffusion
Ziwei Luo, Fredrik K. Gustafsson, Jens Sjölund, Thomas B. Schön
https://tmlr.infinite-conf.org/paper_pages/Eq9k6Va3hY
#diffusion #restoration #stochastic
What is TMLR doing under a 300% increase in submissions?
- Suspensions of >150 authors
- Publicly naming major violations
- Submission quotas distinguishing slop patterns
- Desk rejections 6% → 53%; focus on clarity
- AI reviews evaluating soundness
blog.tmlr.org/2026/adaptin...
- Suspensions of >150 authors
- Publicly naming major violations
- Submission quotas distinguishing slop patterns
- Desk rejections 6% → 53%; focus on clarity
- AI reviews evaluating soundness
blog.tmlr.org/2026/adaptin...
Adapting Our Processes to the Changing World of Scientific Publication
TMLR is adapting its submission and review processes to handle rapid growth while protecting review quality, reviewer time, and scientific integrity.
blog.tmlr.org
October 16, 2026 at 3:13 PM
What is TMLR doing under a 300% increase in submissions?
- Suspensions of >150 authors
- Publicly naming major violations
- Submission quotas distinguishing slop patterns
- Desk rejections 6% → 53%; focus on clarity
- AI reviews evaluating soundness
blog.tmlr.org/2026/adaptin...
- Suspensions of >150 authors
- Publicly naming major violations
- Submission quotas distinguishing slop patterns
- Desk rejections 6% → 53%; focus on clarity
- AI reviews evaluating soundness
blog.tmlr.org/2026/adaptin...
New #TMLR-Paper-with-Video:
EFFEKT: Efficient Federated Knowledge Transfer to Foundation Models
Matteo Caligiuri, Francesco Barbato, Pietro Zanuttigh, Francesco Restuccia
https://tmlr.infinite-conf.org/paper_pages/jpUDUJfE1K
#federated #proxy #privacy
EFFEKT: Efficient Federated Knowledge Transfer to Foundation Models
Matteo Caligiuri, Francesco Barbato, Pietro Zanuttigh, Francesco Restuccia
https://tmlr.infinite-conf.org/paper_pages/jpUDUJfE1K
#federated #proxy #privacy
October 18, 2026 at 4:31 AM
New #TMLR-Paper-with-Video:
EFFEKT: Efficient Federated Knowledge Transfer to Foundation Models
Matteo Caligiuri, Francesco Barbato, Pietro Zanuttigh, Francesco Restuccia
https://tmlr.infinite-conf.org/paper_pages/jpUDUJfE1K
#federated #proxy #privacy
EFFEKT: Efficient Federated Knowledge Transfer to Foundation Models
Matteo Caligiuri, Francesco Barbato, Pietro Zanuttigh, Francesco Restuccia
https://tmlr.infinite-conf.org/paper_pages/jpUDUJfE1K
#federated #proxy #privacy
New #TMLR-Paper-with-Video:
CountLoop: Training-Free High-Instance Image Generation via Iterative Agent Guidance
Anindya Mondal, Sauradip Nag, Ayan Banerjee, Josep Llados, Xiatian Zhu, Anjan Dutta
https://tmlr.infinite-conf.org/paper_pages/2JxXGhpCP4
#scenes #countloop #visual
CountLoop: Training-Free High-Instance Image Generation via Iterative Agent Guidance
Anindya Mondal, Sauradip Nag, Ayan Banerjee, Josep Llados, Xiatian Zhu, Anjan Dutta
https://tmlr.infinite-conf.org/paper_pages/2JxXGhpCP4
#scenes #countloop #visual
October 18, 2026 at 8:32 AM
New #TMLR-Paper-with-Video:
CountLoop: Training-Free High-Instance Image Generation via Iterative Agent Guidance
Anindya Mondal, Sauradip Nag, Ayan Banerjee, Josep Llados, Xiatian Zhu, Anjan Dutta
https://tmlr.infinite-conf.org/paper_pages/2JxXGhpCP4
#scenes #countloop #visual
CountLoop: Training-Free High-Instance Image Generation via Iterative Agent Guidance
Anindya Mondal, Sauradip Nag, Ayan Banerjee, Josep Llados, Xiatian Zhu, Anjan Dutta
https://tmlr.infinite-conf.org/paper_pages/2JxXGhpCP4
#scenes #countloop #visual