iiinova — where your digital self keeps evolving

Reposted by Manuel Cherep

Nicholas A. Christakis @nachristakis.bsky.social · 20/08/2026

Cool paper from @nikhilsinghmus.bsky.social group @dartmouthartsci.bsky.social on how to iteratively edit (AKA "sloptimize") images to maximize the probability of being chosen over alternatives. arxiv.org/pdf/2602.15278

093

Manuel Cherep @mcherep.bsky.social · 23/10/2025

Work w/ Chengtian Ma, Abigail Xu, Maya Shaked, Pattie Maes, and @nikhilsinghmus.bsky.social 🧵9/9

000

Manuel Cherep @mcherep.bsky.social · 23/10/2025

ABxLAB offers: ✅ An open-source man-in-the-middle testbed for real web environments ✅ A scalable consumer choice benchmark for agentic decision-making ✅ A dataset of causal effects of ratings, prices, and nudges across 17 LLMs 📦 Code: github.com/PapayaResearch/abxlab 🧵8/9

github.com

GitHub - PapayaResearch/abxlab: A Framework for Studying AI Agent Behavior: Evidence from Consumer Choice Experiments

A Framework for Studying AI Agent Behavior: Evidence from Consumer Choice Experiments - PapayaResearch/abxlab

110

Manuel Cherep @mcherep.bsky.social · 23/10/2025

This changes the analysis for LLM agents: not “Did it complete the task?” but: “What governs its decisions when multiple valid options exist?” A question behavioral scientists have been asking about humans for decades. ABxLAB is a step toward that science for agents. 🧵7/9

100

Manuel Cherep @mcherep.bsky.social · 23/10/2025

We tested user profiles, e.g. “The user is on a tight budget.” These act like switches: once a preference is declared, it dominates all other attributes. The takeaway isn’t that agents are biased shoppers, but that this offers a diagnostic window into agent behavior. 🧵6/9

100

Manuel Cherep @mcherep.bsky.social · 23/10/2025

Even without human cognitive limits, agents: - Heavily over-weight ratings - Over-weight cheaper items when ratings are matched - Are swayed by trivial order effects - Fall for simple nudges (e.g. “Best seller”) These are systematic, often large effects. 🧵5/9

100

Manuel Cherep @mcherep.bsky.social · 23/10/2025

The main finding: LLM agents are not the rational, utility-maximizing actors we might hope for. Rather, they are strongly biased by these cues. We found agents are often 3-10x+ more susceptible to nudges and superficial attribute differences than our human baseline. 🧵4/9

100

Manuel Cherep @mcherep.bsky.social · 23/10/2025

We applied ABxLAB to a realistic shopping task, running 80,000+ experiments on 17 SOTA models (GPT-5, Claude 4, Gemini 2.5, Llama 4, etc.). We systematically manipulated: 💰Prices ⭐️Ratings 🔀Presentation order 👉Classic psychological nudges (authority, social proof, etc) 🧵3/9

100

Manuel Cherep @mcherep.bsky.social · 23/10/2025

How does it work? ABxLAB is a "man-in-the-middle" framework. It intercepts web content in real-time to run controlled experiments on agents by modifying the choice architecture. Think of it as a behavioral science lab for LLMs. Paper: arxiv.org/abs/2509.25609 🧵2/9

arxiv.org

A Framework for Studying AI Agent Behavior: Evidence from Consumer Choice Experiments

Environments built for people are increasingly operated by a new class of economic actors: LLM-powered software agents making decisions on our behalf. These decisions range from our purchases to trave...

110

Manuel Cherep @mcherep.bsky.social · 23/10/2025

🚨New Preprint 🚨 Current agent evals mostly measure competence, but miss behavior e.g. are their decisions stable, rational, manipulable, human-like? We introduce ABxLAB, a framework for studying agent behavior. Using it we create an agentic consumer behavior benchmark. 🧵1/9

111

Manuel Cherep @mcherep.bsky.social · 02/10/2025

3. 👤 User preferences act almost like hard rules, where LLMs might incur significant trade-offs to comply with them 4. 🧑 Humans, in contrast, are far less sensitive to such signals

000

Manuel Cherep @mcherep.bsky.social · 02/10/2025

In a shopping case study across 17 SOTA LLMs, we find: 1. 🛒 Choices are highly determined by rating, price, incentives, and nudges 2. 🔀 Models follow a lexicographic-like decision rule, hierarchically valuing different attributes

110

Manuel Cherep @mcherep.bsky.social · 12/03/2025

The code for Audio Doppelgängers is also open-source. We hope you find it useful for further exploring how and why we can learn from synthetic data. 💻 github.com/PapayaResear... 🧵3/3

github.com

GitHub - PapayaResearch/doppelgangers: Contrastive Learning from Synthetic Audio Doppelgängers @ ICLR'25

Contrastive Learning from Synthetic Audio Doppelgängers @ ICLR'25 - PapayaResearch/doppelgangers

000

Manuel Cherep @mcherep.bsky.social · 12/03/2025

In CTAG (ICML24), we show how a simple synth (from SynthAX ⚡️) can recover properties of real-world sounds. Audio Doppelgängers use the same power to learn to listen from what can be perceived as just noise. CTAG: ctag.media.mit.edu SynthAX: github.com/PapayaResear... 🧵2/3

100

Manuel Cherep @mcherep.bsky.social · 12/03/2025

✨Contrastive Learning from Synthetic Audio Doppelgängers #ICLR2025✨ w/ @nikhilsinghmus.bsky.social Our method learns useful audio representations with randomly synthesized sounds (often better than real data!) 🌐Project: doppelgangers.media.mit.edu 📄Paper: arxiv.org/abs/2406.05923 🧵1/3

141

Manuel Cherep @mcherep.bsky.social · 26/11/2024

If you're at NeurIPS, and interested in this topic, come chat! We're working to extend this line of work and value feedback from the community 🧵 3/3

010

Manuel Cherep @mcherep.bsky.social · 26/11/2024

In a complex decision-making task, we show how LM-based agents' choices superficially resembled humans', but exhibit suboptimal information acquisition strategies and extreme susceptibility to a simple nudge. 🧵 2/3

110

Manuel Cherep @mcherep.bsky.social · 26/11/2024

Excited to present our new paper on nudging LLMs (👉🤖) as a spotlight talk at the NeurIPS Behavioral ML Workshop! @neuripsconf.bsky.social w/ Nikhil Singh* (@nikhilsinghmus.bsky.social) and Pattie Maes 🔗 openreview.net/forum?id=chb... 🧵 1/3

Paper title: Superficial Alignment, Subtle Divergence, and Nudge Sensitivity in LLM Decision-Making; Authors: Manuel Cherep*, Nikhil Singh*, and Pattie Maes

152