
curr @Covalent_HQ | ex @encifherio @0xStackr @pushchain | @0xpolygon Fellowβ22 | @Arsenal β¦οΈ
Kolkata
Joined January 2021
- Tweets1.7K
- Following747
- Followers1.3K
- Likes2K
Pinned Tweet
Iβve been diving deep into memory engines for LLMs β specifically how far we can reduce token usage while improving long-context accuracy.
So I built a memory layer for LLMs and ran it against the full LongMemEval-S benchmark:
Thanks to @supermemory and @mastra for their benchmarking framework which i used a reference to set up mine
500 questions
gpt-4o actor + gpt-4o judge
Best-of-3 judging passes
Results:
82.22% unweighted accuracy
80.80% micro accuracy
Per-category breakdown:
β’ SSU: 92.86%
β’ SSA: 94.64%
β’ SSP: 70.00%
β’ KU: 88.46%
β’ TR: 76.69%
β’ MS: 70.68%
The system was strongest on single-session factual recall, and weakest on multi-session aggregation + temporal reasoning
While everyoneβs chasing for virality - you chase for performance
Analyse and find videos that would actually help you onboard customers
Try out today at : ugcs.farm/mcp
This Post is from an account that no longer exists. Learn more
2
428
archis.eth π¦π retweeted
Iβve been diving deep into memory engines for LLMs β specifically how far we can reduce token usage while improving long-context accuracy.
So I built a memory layer for LLMs and ran it against the full LongMemEval-S benchmark:
Thanks to @supermemory and @mastra for their benchmarking framework which i used a reference to set up mine
500 questions
gpt-4o actor + gpt-4o judge
Best-of-3 judging passes
Results:
82.22% unweighted accuracy
80.80% micro accuracy
Per-category breakdown:
β’ SSU: 92.86%
β’ SSA: 94.64%
β’ SSP: 70.00%
β’ KU: 88.46%
β’ TR: 76.69%
β’ MS: 70.68%
The system was strongest on single-session factual recall, and weakest on multi-session aggregation + temporal reasoning
The more interesting part was token efficiency.
Compared to the full-history Chain-of-Note baseline, on the same 500-question dataset and prompt template:
β’ ~3,200 input tokens/question vs ~14,600
β’ 78% fewer input tokens
β’ +22pp higher accuracy
1
1
110
The architecture ended up being a mix of retrieval, reranking, and fallback reasoning:
Retrieval side:
β’ Hybrid dense + BM25 fusion
β’ Heuristic + cached LLM query expansion
β’ MMR diversity
β’ Recency bias for knowledge-update questions
β’ Time-anchor bias for temporal questions
β’ Optional cross-encoder + LLM listwise rerank
Answer side:
β’ Chunk-level hits rolled up into session-level evidence
β’ Retrieval-confidence scoring
β’ Low-confidence fallback that injects raw transcript from top sessions
β’ Pluggable single-shot or self-critique synthesizer
2
96
archis.eth π¦π retweeted
Hyperliquid Markets Weekly Heatmap
DeFi: +11.4% β
Stocks: +11.2% β
Layer 1: +4.0% β
HIP-3: +3.8% β
Indices: +3.3% β
Meanwhile, AI: -17.5%. RWA: -5.3%. Commodities: -4.5%.
DeFi and Stocks running together while AI gets wiped
The market isn't rotating out of crypto. It's rotating out of the narrative - How are you positioning yourself?
Source: hyperliquid.goldrush.dev
archis.eth π¦π retweeted
PRIVATE BETA LIVE
Catch what others miss, before itβs too late
Trench is here. Lock in, dominate the market π¦ΎπΌ
archis.eth π¦π retweeted
CompliantLLM (@FiddleCubeAI) detects data leaks into any third-party GenAI tools used across your company. It identifies GenAI-specific attacks to catch breaches in both approved and unapproved AI workflows.
ycombinator.com/launches/OBoβ¦
Congrats on the launch, @kaushik_himself & @nupoor_neha
archis.eth π¦π retweeted
Trenchers, stand by!
A silent vessel cuts through the dark, armed with what comes next
[PRIVATE BETA] is loadingβ¦
archis.eth π¦π retweeted
I can literally buy TSLA stock using TRUMP memecoin XD
That too even privately
Using @encifherio Thanks to @JupiterExchange
Played around with @coinbase x402 protocol and hereβs what I built recently
Codey β¨ - unlock true pay-as-you-go in your AI powered code editors. Use your wallet as API key, top-up USDC & pay-per-call. Pick your context, let smart agents pick the best model for price & accuracy. You control the budget, no flat subscriptions, no wastage.
Where to next
- Currently wallet is stored in VSCode SecretStorage, can be extended to use CDP wallet with session keys and permissions
- Democratising model choice, anyone can host their service and provide api endpoints and set their fees
Check out the demo, DM to get access and try out
cc @CoinbaseDev
archis.eth π¦π retweeted
π We've raised $4.2M to modernize financial operations for crypto companies!
Led by @1kxnetwork with @ycombinator, @GoodwaterCap, @DCGco, and @SevenXVentures
Just like Brex transformed traditional business finance, we're making crypto ops seamless.
archis.eth π¦π retweeted
Introducing: dsheets.new π©βπ§
Decentralized spreadsheets made for the global economy & open Internet
archis.eth π¦π retweeted
Give account recovery using email guardian a try -> recovery.zk.email
Use passkey for a smooth experience π
We keep pushing Account Recovery features - and we have a new demo to show to you using Passkeys and EIP-7702 π¬
Delivered by: @0xdhruva π












