Tejas Vaidhya (@imtejas13) profile picture

Building @nolanoorg | Curious about everything | Previously: @Mila_Quebec, @iitkgp, @cern

San Francisco, CA
Joined April 2020
Photo by Tejas Vaidhya (@imtejas13)Photo by Tejas Vaidhya (@imtejas13)Photo by Tejas Vaidhya (@imtejas13)Photo by Tejas Vaidhya (@imtejas13)Photo by Tejas Vaidhya (@imtejas13)Photo by Tejas Vaidhya (@imtejas13)Photo by Tejas Vaidhya (@imtejas13)Photo by Tejas Vaidhya (@imtejas13)Photo by Tejas Vaidhya (@imtejas13)
Tejas Vaidhya retweeted
Bonjour, Montréal.
173
94
11
1,406
70,817
Tejas Vaidhya retweeted
Replying to @neuralink
Really excited to see this progress! We’ve been exploring some of these ideas in our work on neuro-foundation models. Glad to see these models enabling live BCI use.
This new post from @neuralink combines POYO-style spike tokenisation with a Mamba backbone, akin to our POSSM architecture, and explores self-supervised learning for these models, as we did recently with MOJO. Exciting to see these ideas supporting real-time BCI use! 🧵
3
10
495
Tejas Vaidhya retweeted
"It's the best money I spent whole year" is not a compliment I ever imagined we'd get from a person who's tried and used every AI tool imaginable Our team spends enormous effort (and time!) in making Slidely feel BAU for our users, and hearing things like these makes it all worth it❤️
1
4
3
16
1,013
Tejas Vaidhya retweeted
brought together a bunch of logistics and supply chain folks to experience what post singularity work would look like: dinner, drinks and dealmaking
2
1
1
20
850
Tejas Vaidhya retweeted
BREAKING: Kalpa TTS Beta v0.1 by @KalpaLabsAI takes #2 on Audio Realism Bench with an Elo of 1331. The model ranks just ahead of MAI-Voice-2 by @MicrosoftAI and Eleven v3 Conversational by @ElevenLabs, and behind Bland Speech v3 by @usebland. Congratulations to @pshishodiaa, @gautam_jha_37, and the @KalpaLabsAI team!
11
21
8
201
31,646
Today, we're releasing our conversational speech model in public beta, and sharing our broader vision for Generalist Audio Models (GAMs): unified audio models that can follow complex instructions and learn from in-context examples.
14
31
13
108
14,070
Tejas Vaidhya retweeted
hello there the jacobian conjecture is false thanx to my close friend akhil for asking about it and my other close friend fable for working during the world cup final ((1+xy)^3 z + y^2 (1+xy) (4+3xy), y + 3 x (1+xy)^2 z + 3 x y^2 (4+3xy), 2 x - 3 x^2 y - x^3 z): \C^3\to \C^3, has jacobian determinant -2, and sends (0, 0, -1/4), (1, -3/2, 13/2), and (-1, 3/2, 13/2) to (-1/4, 0, 0)
1,698
5,362
3,156
43,834
40,132,921
Tejas Vaidhya retweeted
Replying to @satyanadella
@satyanadella’s post today is about where durable value gets created - a learning organisation where context compounds. This aligns with how I’ve been thinking about what it’s going to take to build the enterprise context layer. xitter.gq/prukalpa/status/206151…
1
2
19
1,986
Tejas Vaidhya retweeted
Swish has raised $38M to bring fresh food to more neighbourhoods 💚 Led by Hara Global, @BainCapVC, and @AccelIndia. Started 18 months back and here we are. @aniketshah30 @saranonearth
22
13
2
213
11,254
Tejas Vaidhya retweeted
Kalpa Labs (@KalpaLabsAI) is training one generalist model for all speech tasks. Steer it like an LLM, give long system prompts & make use of native contextual awareness. Congrats on the launch, @pshishodiaa and @jhagauti! ycombinator.com/launches/Op4…
22
35
13
268
97,489
Fundraise Announcement! We are proud to announce that we at @smallest_AI have raised 8M USD in an oversubscribed seed round led by @Sierra_Ventures with participation from @3one4Capital, @better_capital, @UpsparksCapital, @SchemaVentures, @DeVC_Global, @tinyvc, @peercheque, @shyamalanadkat, @missionstcap, @bwertz, @RaveenSastry, and many more angels to build the future of Enterprise Voice AI! We are transforming voice AI from the ground up - pushing agents to pass the Turing test in the coming years. We are also excited to build a research-first organization - diving into the depths of problems from first principles and doing things no one in the world is even thinking about. It was our dream to build a scientific team that pushes the boundaries of the frontier. We are privileged to have the support of some of the best early backers a founder could ever ask for. Thank you, everyone for supporting us - this is just the beginning!
211
95
40
1,500
680,510
Congratulations to Irina Rish (@irinarish) Core Academic Member at Mila, on being named a Principal Investigator for the @SimonsFdn's new Simons Collaboration on the Physics of Learning and Neural Computation. This initiative brings together researchers from diverse disciplines to deepen our understanding of how large neural networks learn, reason, and imagine, essentially peeking inside AI’s “blackbox." simonsfoundation.org/2025/08…
1
4
27
3,973
Tejas Vaidhya retweeted
Great work by my amazing collaborators on memory- and inference-efficient quantized LLMs! ACL paper presents efficient inference kernels for our Spectra models introduced earlier in ICLR paper.
🚀 New Paper: Scaling Laws and Efficient Inference for Ternary Language Models. Thrilled to share that our work was presented at ACL 2025! We explore ternary LMs (TriLMs), studying their scaling laws and efficiency compared to traditional FloatLMs. 🧵 1/6
6
18
1,529
Inspired by “minimal implementation“ projects in AI such as @karpathy’s nanoGPT, I worked to bring this concept to the HPC world! I’ve built a minimal implementation of an MPI library called nanoMPI, which focuses on clarity, simplicity, and easy installation.
12
36
2
288
38,455
Congratulations to @Yoshua_Bengio on receiving the prestigious King Charles III Coronation Medal in recognition of his contributions to the field of deep learning!
On Feb. 28, I was honoured to receive the King Charles III Coronation Medal in recognition of my contributions to deep learning. Thank you to Senator Andrew Cardozo for presenting me with this important distinction!
2
1
44
3,139
Learned optimizers can’t generalize to large unseen tasks…. Until now! Excited to present μLO: Compute-Efficient Meta-Generalization of Learned Optimizers! Don’t miss my talk about it next Sunday at the OPT2024 Neurips Workshop :) 🧵arxiv.org/abs/2406.00153 1/N
2
33
5
112
12,827
Tejas Vaidhya retweeted
Replying to @BensenHsu @sama
Indeed, memory is becoming a bottleneck, but there is lots of work on quantization/compression and also training low-bitwidth models from scratch works well (shameless plug to our TriLMs /Spectra scaling suite w/ @imtejas13 et al at @NolanoOrg : arxiv.org/abs/2407.12327
1
2
4
705
Tejas Vaidhya retweeted
Looking for PhD interns for Summer 2025 with strong publication records. Experience with video-language foundation models is a plus. US-based role with the chance to publish from the internship. If interested or can recommend a candidate, please reach out. jobs.apple.com/en-us/details…
24
63
6
444
183,296
Spectra by @NolanoOrg is an open suite of 54 LLMs spanning FP16 training, ternary training, and post-training quantisation to 3/4/6/8 bits. We're excited to see momentum and great open science around ternary models! @_AyushKaushal @imtejas13 @irinarish xitter.gq/NolanoOrg/status/18139…
🚀 SpectraSuite of Ternary and FP16 LLMs 🚀 We’re thrilled to release the Spectra Suite of open ternary (TriLMs) and FP16 (FloatLMs) language models from 99M to 3.9B parameters. At billion+ parameter scale, TriLMs upto 10x smaller can match the performance of FloatLMs. 1/5
1
2
2
894
Very cool paper that shows impressive performance with ternary LLMs. Discovering new papers that use @AiEleuther's GPT-NeoX library in the wild is always a treat as well :D
🚀 SpectraSuite of Ternary and FP16 LLMs 🚀 We’re thrilled to release the Spectra Suite of open ternary (TriLMs) and FP16 (FloatLMs) language models from 99M to 3.9B parameters. At billion+ parameter scale, TriLMs upto 10x smaller can match the performance of FloatLMs. 1/5
1
3
17
3,175