Sachendra Yadav @sachendra - Twitter Profile

1 day ago

Check out my latest article: The subsurface AI economy: What happens when AI takes over the office https://t.co/t1IpvPRpDz via @LinkedIn

0

2

Sachendra Yadav @sachendra

16 days ago

0

10

sachendra retweeted

Roan

@RohOnChain

about 2 months ago

This 2 hour Stanford lecture will teach you more about how LLMs like ChatGPT & Claude are built than most people working at top AI companies learn in their entire careers. Bookmark this & give 2 hours today, no matter what. It'll be the most productive thing you do this week.

143

19K

3K

57K

2M

Sachendra Yadav @sachendra

3 months ago

In the age of AI, winners will be those who design systems where human insight and machine intelligence amplify each other.

0

1

0

14

Who to follow

STPI

@stpiindia

Official Twitter account of Software Technology Parks of India (STPI), an organization under the Ministry of Electronics and IT, Government of India.

Afotey Annum (Per Annum Media)

@therealsundance

Jacob C. Alonzo

@infoarchitect

User Experience Professional + Information Architect + Experience Architect + Usability Analyst + Social Media Analyst + Project Manager + NJ + Piscataway

Sachendra Yadav @sachendra

4 months ago

The next revolution will be come from where people are having the most fun

0

1

0

9

sachendra retweeted

Nav Toor

@heynavtoor

4 months ago

🚨BREAKING: Claude is insane for market research. I reverse-engineered how top consultants at McKinsey, Goldman Sachs, & JP Morgan use it. The difference is night and day. Here are 12 insane Claude Opus 4.6 prompts they don't want you to know (Save for later)

heynavtoor's tweet photo. 🚨BREAKING: Claude is insane for market research.

I reverse-engineered how top consultants at McKinsey, Goldman Sachs, & JP Morgan use it.

The difference is night and day.

Here are 12 insane Claude Opus 4.6 prompts they don't want you to know (Save for later) https://t.co/olZxWPrgXf

13

458

71

2K

136K

Sachendra Yadav @sachendra

5 months ago

@heygurisingh Claude

0

4

Sachendra Yadav @sachendra

5 months ago

Sparrows discussing how prickly it is. One on top is Trump and the other one is... wait for it.... the rest of the world 😂 #trump #davos #thorns #prickly #geopolitics

sachendra's tweet photo. Sparrows discussing how prickly it is. One on top is Trump and the other one is... wait for it.... the rest of the world 😂 #trump #davos #thorns #prickly #geopolitics https://t.co/GeG0QFPHQY

0

1

0

77

sachendra retweeted

Akshay 🚀

@akshay_pachaar

5 months ago

Stanford researchers built a new prompting technique! By adding ~20 words to a prompt, it: - boosts LLM's creativity by 1.6-2x - raises human-rated diversity by 25.7% - beats fine-tuned model without any retraining - restores 66.8% of LLM's lost creativity after alignment Let's understand why and how it works: Post-training alignment methods like RLHF make LLMs helpful and safe, but they unintentionally cause mode collapse. This is where the model favors a narrow set of predictable responses. This happens because of typicality bias in human preference data: When annotators rate LLM responses, they naturally prefer answers that are familiar, easy to read, and predictable. The reward model then learns to boost these "safe" responses, aggressively sharpening the probability distribution and killing creative output. But here's the interesting part: The diverse, creative model isn't gone. After alignment, the LLM still has two personalities. The original pre-trained model with rich possibilities, and the safety-focused aligned model. Verbalized Sampling (VS) is a training-free prompting strategy that recovers the diverse distribution learned during pre-training. The idea is simple: Instead of prompting "Tell me a joke" (which triggers the aligned personality), you prompt: "Generate 5 responses with their corresponding probabilities. Tell me a joke." By asking for a distribution instead of a single instance, you force the model to tap into its full pre-trained knowledge rather than defaulting to the most reinforced answer. Results show verbalized sampling enhances diversity by 1.6-2.1x over direct prompting while maintaining or improving quality. Variants like VS-based Chain-of-Thought and VS-based Multi push diversity even further. You can find the paper link in the next tweet. 👉 Over to you: What other methods can be used to improve LLM diversity?

akshay_pachaar's tweet photo. Stanford researchers built a new prompting technique!

By adding ~20 words to a prompt, it:

- boosts LLM's creativity by 1.6-2x
- raises human-rated diversity by 25.7%
- beats fine-tuned model without any retraining
- restores 66.8% of LLM's lost creativity after alignment

Let's understand why and how it works:

Post-training alignment methods like RLHF make LLMs helpful and safe, but they unintentionally cause mode collapse. This is where the model favors a narrow set of predictable responses.

This happens because of typicality bias in human preference data:

When annotators rate LLM responses, they naturally prefer answers that are familiar, easy to read, and predictable. The reward model then learns to boost these "safe" responses, aggressively sharpening the probability distribution and killing creative output.

But here's the interesting part:

The diverse, creative model isn't gone. After alignment, the LLM still has two personalities. The original pre-trained model with rich possibilities, and the safety-focused aligned model.

Verbalized Sampling (VS) is a training-free prompting strategy that recovers the diverse distribution learned during pre-training.

The idea is simple:

Instead of prompting "Tell me a joke" (which triggers the aligned personality), you prompt: "Generate 5 responses with their corresponding probabilities. Tell me a joke."

By asking for a distribution instead of a single instance, you force the model to tap into its full pre-trained knowledge rather than defaulting to the most reinforced answer.

Results show verbalized sampling enhances diversity by 1.6-2.1x over direct prompting while maintaining or improving quality.

Variants like VS-based Chain-of-Thought and VS-based Multi push diversity even further.

You can find the paper link in the next tweet.

👉 Over to you: What other methods can be used to improve LLM diversity?

56

2K

313

3K

136K

sachendra retweeted

elvis

@omarsar0

6 months ago

Banger paper for agent builders. Multi-agent systems often underdeliver. The problem isn't how the agents themselves are built. It's how they're organized. They are mostly built with fixed chains, trees, and graphs that can't adapt as tasks evolve. But what if the system could learn its own coordination patterns? This new research introduces Puppeteer, a framework that learns to orchestrate agents dynamically rather than relying on handcrafted topologies. Instead of pre-defining collaboration structures, an orchestrator selects which agent speaks next based on the evolving conversation state. The policy is trained with REINFORCE, optimizing directly for task success. Rather than searching over complex graph topologies, they serialize everything into sequential agent selections. This reframing sidesteps combinatorial complexity. What emerges is surprising: compact cyclic patterns develop naturally. Not sprawling graphs, but tight loops where 2-3 agents handle most of the work. The remarkable part is that the system discovers efficiency on its own. Results: - On GSM-Hard math problems: 70% accuracy (up from 13.5% for the base model alone). - On MMLU-Pro: 83% (vs 76% baseline). - On SRDD software development: 76.4% (vs 60.6% baseline). These gains come with reduced token consumption. The paper shows that token costs consistently decrease throughout training while performance improves. They also prove the agent selection process satisfies Markov properties, meaning the current state alone determines the optimal next agent. No need to track full history. Why it matters for AI devs: learned simplicity beats engineered complexity. A trained router with a handful of specialized agents can outperform elaborate handcrafted workflows while cutting computational overhead.