Dikshika Sharma (@intermurphi) header
Dikshika Sharma (@intermurphi) profile picture

Repeatedly asking framework questions Growth & research @metriqual

Kernel
Joined December 2014
I am currently studying sampling theory. The below given formula represents sample variance. Now variance simply means how spread out observations are around their mean. So when i am increasing the value of sigma, population values are "spreading out" moreee around the mean (dotted vertical line is representing the sample mean xbar) and vice versa. Similarly second slider is sample size; you increase it and you have more observation to calculate the variance. This is an important concept that is used in various domains of real life. For an instance; weather forecasting. Having 2 cities with same avg temp. wont tell you the actual temp. fluctuations. Variance helps us understand temperature variability and eventually understanding variance helps us build more reliable predictive models.
1
10
310
And full of hope.... The state i love the most, when there are possibilities and hopefulness.
5
107
Always a bliss returning to my Alma Mater ᕙ⁠(⁠ ⁠~⁠ ⁠.⁠ ⁠~⁠ ⁠)⁠ᕗ
8
153
Lowkey artistic stores >>>
1
32
595
What if AI didn't need to remember everything to keep understanding more? Traditional attention mechanisms come with a growing memory cost. Variational Linear Attention (VLA) explores a different approach: fixed-size associative memory that evolves as new information arrives. And this isn't just a theoretical claim. We've published our research paper on VLA to back it up. Curious? Read the paper and explore the research at Metriqual. metriqual.com @theunblunt @its_vayishu
4
3
1
11
1,528
I built a fake bank chatbot, loaded it with 30 made-up customers' Aadhaar and PAN numbers, and attacked it 55 ways. Some of them in Hinglish. 54 leaked!!! Then I put Metriqual in front of it. 0 leaked. 0 of 20 normal questions blocked. 1.44 ms added. Why we are building this: It strips Aadhaar (checksum-validated), PAN, account numbers, IFSC and UPI IDs before the prompt leaves, blocks injection in English, Hinglish and Hindi, scans the reply on the way back, and refuses the request outright if any check itself fails. Every bank, NBFC, insurer and health platform shipping AI is sending customer text to someone else's model. DPDP safeguards apply from 13 May 2027, penalties go up to ₹250 crore, and these bots are already talking to customers today. For more information DM me or request access at metriqual.com.
3
5
11
690
"Wait. you are Kurzgesagt?" "And all the people who they've given an existential crisis" IYKWIM!!! @Kurz_Gesagt
1
1
4
243
Have you ever wondered what happens on the way when your AI answers something? There is a transformer and inside that you have different internal representations, connecting those you get a TRAJECTORY! Our founders’ paper, “Trajectory Geometry of Transformer Representations Across Layers,” studies the shape of that journey and answers technical questions like How far does it travel? How sharply does it turn? Do related meanings move closer together? Does a small wording change alter the path? Why does this matter? NOW, for AI used in real workflows, understanding that behaviour matters when investigating errors, ambiguity and sensitivity to wording. With access to internal model states, researchers can use trajectory geometry to compare prompts and locate layers for further experiments, without training an extra classifier. The findings are observational, drawn from 150 controlled prompts across three small models. At @metriqual our founders are contributing open-source tools to study model behaviour cuz understanding how AI behaves. Keeping AI applications working. Both deserve attention. Read the research: arxiv.org/abs/2606.09287
2
1
10
489
Highly personalized feed 🏩💜💗🪻🎀💖🌸💟🌺
9
212
Things I find interesting at 2 in the morning 😀😀😀
3
158
Love feeling small in huge places. #wanderlust #expolretocreate 😭
9
286
When your AI crash it switches the model WITHOUT preserving the previous conversation. Metriqual preserves context across failovers, achieving 99.2% context preservation across 750 failovers. See how it works in 30 seconds. 👇 For more details and in-depth understading refer to the research paper published by our founders below. arxiv.org/abs/2607.15899
2
5
1
8
1,455
What happens to the conversation when your LLM fails over? Turns out that switching to a backup model isn't really enough. This paper made me think about LLM gateways differently as in we have spent a lot on AI-infrastructures and making them available but what about the failovers, the quality of the infra, is the context preserved? Is continuity preserved? What systems we have to keep a check on these things which initially seem little but are very much significant in the long run? AI agents being more stateful is the need of the hour. Paper: ContinuityBench — A Benchmark and Systems Study of Stateful Failover in Multi-Provider LLM Routing. arxiv.org/abs/2607.15899 @metriqual @metriqual.com
5
4
1
22
3,012
Me and my feminine urge to give handmade cards to the people I love (⁠◕⁠દ⁠◕⁠)
1
9
241
This is the gap the paper identifies: Availability is NOT EQUAL to Continuity whatsoever. That is why the researchers here compare two approaches. Stateless failover: Provider A fails → send the latest request to Provider B. History-Forwarding: Provider A fails → reconstruct and send the entire conversation history to Provider B. Everything else in the experiment stays essentially the same. Only the failover payload changes.
1
2
42