#resnet search results
Getting a shape mismatch error when using resnet18 with Fashion MNIST? ResNet expects 3-channel (RGB) images, but FMNIST is grayscale (1-channel). #computervision #resnet
#CLIP #ResNet RN50x4 #deepdream @ penultimate layer ooooooo horrifying pug'o'pizzapeople python programmer's despair #AI #AIart #neuron #dreams #polysemantic #multimodal
#CLIP #ResNet RN50x4 #deepdream Layer1 Layer 1 maximum dog feature abstraction #AI #AIart #neuron #dreams
distorSurreal horrifying zoKaleidoPhobia demonic PsychedPsychOPaware psychedUFO #CLIP #ViT reproducing a #ResNet neuron in its neural net #AIart #gradientascent #modelinversion
ResNet's skip connections solve the degradation problem in very deep networks. Information highways allow gradients to flow unimpeded to early layers. Sometimes the best path forward is a shortcut. #DeepNeuralNetworks #ResNet #SkipConnections andrewroche.ai/deep-neural-ne…
#CLIP #ResNet RN50x4 #deepdream Gradient got hijacked by burnt-in text: ARF AFF WOOOOF WWWW OOOO. 🤣 Layer2+Layer3 combo effort. Probably needs normalization. #AI #AIart #neuron #dreams
#CLIP #ResNet RN50x4 #deepdream Layer2 How does a CLIP see little dog faces in a dog? - As many many many faces. And as TEXT. Of course. 🙃 #AI #AIart #neuron #dreams #scribble #doodle
Great to joining @NIHRSPHR #ResNet panel on research and the media with colleagues @spidermaani @marcobardus @PaulineScheelb1
Not sure if an autoencoder's loss is a valid measure of 'crappiness of text encoder' (especially re: dims!) as it tries to learn #CLIP ViT-L/14 text embeddings [40k labels] <-> #ResNet embeds --VS-- #T5 embeds, but - this is probably not good. 😂 Though it IS interesting! ☝️🤓
![zer0int1's tweet image. Not sure if an autoencoder's loss is a valid measure of 'crappiness of text encoder' (especially re: dims!) as it tries to learn #CLIP ViT-L/14 text embeddings [40k labels] &lt;-&gt; #ResNet embeds --VS-- #T5 embeds, but - this is probably not good. 😂
Though it IS interesting! ☝️🤓](https://pbs.twimg.com/media/Gc5F0N7aAAAPkq-.png)
#CLIP #ResNet RN50x4 #deepdream @ penultimate layer My personal winner of category: "ubermultimodal madcap megacomposite" (within top_k=20 most-salient AI self-selected #image #abominations): #AI #AIart #neuron #dreams #polysemantic #multimodal
#CLIP #ResNet RN50x4 #deepdream @ penultimate layer Very awkward way to make a dog. Once you see it... 🙈 Choice for salience made by AI itself (via activation value). Probably as a polysemantic concept. -> Synonym for 'female dog' = 'itch' that starts with a 'B'. #AI #AIart
#CLIP #ResNet RN50x4 #deepdream @ penultimate layer Faces. Knitting. Knitted pugfaces. The visually closest to the original I could find within AI's self-selected most salient top_k=20 (by activation value). #AI #AIart #neuron #polysemantic #multimodal #dreams
#CLIP #ResNet RN50x4 #deepdream @ penultimate layer 🚗: has freak accident 🤖: sees face in car like a human would, alas puts eye in headlight prominently 🤖: adds many many more faces and tiny night city scenes, because AI also saw wall / building #AI #AIart #neuron #dreams
Happy Tuesday! Grateful for this nomination & appreciation of our son. Full time student. 2 jobs. Proud of you Brady❤️ @AugustWest420 @b_callahan8 2 years down 🙏🏼 2 to go #BSU #IT #ResNet #ComputerScience #studentemployee
This is a very cool paper. Decided to take some stabs at simple related circuits. The Impact of Different Activation Functions—ReLU, Swish, and Mish—on the Performance and Feature Diversity of Deep #Neural Networks, Specifically #ResNet Models 💻💻💻 This experiment is…
What “feature diversity” means here is neighboring layers should NOT do almost the same thing – that’s a waste of capacity! New paper: arxiv.org/abs/2310.02244 I cannot show enough appreciation to my coauthors @dingli_yu @chenzhucs @hayou_soufiane !!! More explanations 👇
ResNet's skip connections solve the degradation problem in very deep networks. Information highways allow gradients to flow unimpeded to early layers. Sometimes the best path forward is a shortcut. #DeepNeuralNetworks #ResNet #SkipConnections andrewroche.ai/deep-neural-ne…
Highlight AGAIN: We should understand the role of #residuals from the #representation view, not the gradient vanishing. The authors of #ResNet didn’t understand what causes performance #degradation.
Ilya: LSTM is a variant of ResNet. This is also wrong. The reasons are as follows: LSTM is used to solve the gradient vanishing problem, while residual connection is to reduce the loss of information between layers (of course, the multi-branch structure in LSTM also reduces it).
ResNet explained: Residual connections (skip layers) help networks bypass transformations so gradients flow better. This lets models go much deeper without degradation. Read more info : nomidl.com/natural-langua… #ResNet #DeepLearning #AI #MachineLearning
🔥 Read our Highly Cited Paper 📚#Methodology for Evaluating the Generalization of ResNet 🔗mdpi.com/2076-3417/14/9… 👨🔬by Anan Du et al. @CAS__Science #generalizationevaluation #ResNet
ICYMI: On the latest #RESTalk, Kevin Smith, CEO of @HabitatNCC discusses the groundbreaking ENERGY STAR and DOE Zero Energy Ready Home-certified Bennett Point project in DE. Listen at restalk.libsyn.com/ep143-sweat-eq… #HabitatForHumanity #RESNET #ZeroEnergyHome #ENERGYSTARHomes
Laurel Elam and Ryan Meres of RESNET at Clayton Homes' building facility in TN, learning about their eBuilt® homes! #RESNET #ClaytonHomes #eBuilt #ZERH
Kevin Smith, CEO of @HabitatNCC discusses the groundbreaking project, ENERGY STAR and DOE Zero Energy Ready Home-certified Bennett Point in DE Listen at ow.ly/h8KM50WCPuh #HabitatForHumanity #RESNET #ZeroEnergyHome #EnergyStarHomes
Join RESNET® at the Sunbelt Builders Show in Texas! Visit booth 615 to see why 2 out of 3 new homes built last year earned a HERS Index Score! #RESNET #HERS #sunbelt2025 #WeBuildTexas #EverythingsBiggerinTX
When to Use ResNet in Machine Learning (Simple Decision Guide!) medium.com/nextgenllm/the… #MachineLearning #DeepLearning #ResNet #AI #DataScience #TechTips
Tune in for the annual HERS Trends Report! Hear from Ryan Meres on 2024's biggest HERS Score increases, and trends in foundations, air exchanges, and water efficiency. restalk.libsyn.com/ep142-home-eff… #RESTalk #TrendsinHERS #RESNET #BuildingScience #HomePerformance
🔥 Read our Paper 📚 #Building Ensemble of #Resnet for #Dolphin #Whistle Detection 🔗 mdpi.com/2076-3417/13/1… 👨🔬 by Prof. Loris Nanni et al. 🏫 @UniPD_Research / @MissouriState #CNN
Pushing the Limits of Sparsity: A Bag of Tricks for Extreme Pruning openreview.net/forum?id=XX9Jd… #sparse #imagenet #resnet
#CLIP #ResNet RN50x4 #deepdream @ penultimate layer ooooooo horrifying pug'o'pizzapeople python programmer's despair #AI #AIart #neuron #dreams #polysemantic #multimodal
#CLIP #ResNet RN50x4 #deepdream Layer1 Layer 1 maximum dog feature abstraction #AI #AIart #neuron #dreams
Getting a shape mismatch error when using resnet18 with Fashion MNIST? ResNet expects 3-channel (RGB) images, but FMNIST is grayscale (1-channel). #computervision #resnet
distorSurreal horrifying zoKaleidoPhobia demonic PsychedPsychOPaware psychedUFO #CLIP #ViT reproducing a #ResNet neuron in its neural net #AIart #gradientascent #modelinversion
#CLIP #ResNet RN50x4 #deepdream Layer2 How does a CLIP see little dog faces in a dog? - As many many many faces. And as TEXT. Of course. 🙃 #AI #AIart #neuron #dreams #scribble #doodle
Big #ResNet #CLIP model after #CLIPSurgery, opinion: 🤖: dumb horny dementidata 🤖: ❤️ isnt bananregulatory boring 🤖: banandeeplearning jailbreak ????????? 🤔🤔🤔 🤣 #AI #hilarious #vision #transformer #lol #meme #emoji #prediction
#CLIP #ResNet RN50x4 #deepdream Gradient got hijacked by burnt-in text: ARF AFF WOOOOF WWWW OOOO. 🤣 Layer2+Layer3 combo effort. Probably needs normalization. #AI #AIart #neuron #dreams
#CLIP #ResNet RN50x4 #deepdream @ penultimate layer My personal winner of category: "ubermultimodal madcap megacomposite" (within top_k=20 most-salient AI self-selected #image #abominations): #AI #AIart #neuron #dreams #polysemantic #multimodal
Not sure if an autoencoder's loss is a valid measure of 'crappiness of text encoder' (especially re: dims!) as it tries to learn #CLIP ViT-L/14 text embeddings [40k labels] <-> #ResNet embeds --VS-- #T5 embeds, but - this is probably not good. 😂 Though it IS interesting! ☝️🤓
#CLIP #ResNet RN50x4 #deepdream @ penultimate layer Very awkward way to make a dog. Once you see it... 🙈 Choice for salience made by AI itself (via activation value). Probably as a polysemantic concept. -> Synonym for 'female dog' = 'itch' that starts with a 'B'. #AI #AIart
#CLIP #ResNet RN50x4 #deepdream @ penultimate layer Faces. Knitting. Knitted pugfaces. The visually closest to the original I could find within AI's self-selected most salient top_k=20 (by activation value). #AI #AIart #neuron #polysemantic #multimodal #dreams
#CLIP #ResNet RN50x4 #deepdream @ penultimate layer 🚗: has freak accident 🤖: sees face in car like a human would, alas puts eye in headlight prominently 🤖: adds many many more faces and tiny night city scenes, because AI also saw wall / building #AI #AIart #neuron #dreams
In San Diego for RESNET? Visit us at Table 14 to learn about our high performance spray foam insulation solutions and how they enhance the building envelope. #HuntsmanBuildingSolutions #Huntsman #RESNET #RESNET2023
Oh so that's why this little #CLIP #ResNet #shoggoth #AI just won't get smaller and more efficient. It's just bumming around in a float32 because it has to.🤡 The CLIP #Vision #Transformer, on the other hand, just loves matmulling about an issue on the Tensor cores.🤖 Fun! 🤓
A hybrid model using both ResNet and LSTM to improve accuracy in stock price predictions. #StockPriceForecasting #ResNet #LSTM
ResNLS combines ResNet and LSTM for superior stock price prediction by focusing on price dependencies. #AI #StockPriceForecasting #ResNet
With a distributed #architecture that maximizes #powerefficiency, Hailo-8 beats @Nvidia on #ResNet 50 benchmark. #edgeAI #AI #accelerators #AIchips @Hailo_ai @DanonOrr @JagdaleSaumitra eetimes.eu/hailo-expands-…
Something went wrong.
Something went wrong.
United States Trends
- 1. Veterans Day 35K posts
- 2. Packers 101K posts
- 3. Good Tuesday 21.5K posts
- 4. SoftBank 4,653 posts
- 5. Eagles 129K posts
- 6. Benítez 14.1K posts
- 7. #WWERaw 142K posts
- 8. Jordan Love 15.6K posts
- 9. Bond 59.9K posts
- 10. Matt Lafleur 9,081 posts
- 11. JOONGDUNK BRIGHT SKIN 273K posts
- 12. Green Bay 19.2K posts
- 13. Jalen 24.3K posts
- 14. Berkeley 72.4K posts
- 15. Sirianni 5,217 posts
- 16. Kevin Patullo 7,182 posts
- 17. AJ Brown 7,336 posts
- 18. 600,000 Chinese 7,855 posts
- 19. Tsushima 2,380 posts
- 20. McManus 4,563 posts