Skip to main content

Buildability / Receipt

Do LLMs Need to See Everything? A Benchmark and Study of Failures in LLM-driven Smartphone Automation using Screentext vs. Screenshots

This public receipt window renders only fields present in the canonical receipt object, deterministic fixture receipt, or canonical evidence receipt. Missing compute, demo, hash, signature, approval, telemetry, and adoption fields stay explicit.

Public buildability page receipt window

Watch and verify: Do LLMs Need to See Everything? A Benchmark and Study of Failures in LLM-driven Smartphone Automation using Screentext vs. Screenshots

/buildability/do-llms-need-to-see-everything-a-benchmark-and-study-of-failures-in-llm-driven-smartphone-automation-using-screentext-vs

Watchwatch

Subject: Do LLMs Need to See Everything? A Benchmark and Study of Failures in LLM-driven Smartphone Automation using Screentext vs. Screenshots

Verdict

Watch

Verdict is Watch because viability or proof quality is intermediate and should be re-evaluated before execution.

Time to first demo

Insufficient data

No first-demo timestamp, owner estimate, or elapsed demo receipt is attached to this surface.

Compute envelope

Data

{"file name": "input.pdf", "number of pages": 29, "author": "Shiquan Zhang; Tianyi Zhang; Le Fang; Simon D'Alfonso; Hong Jia; Vassilis Kostakos"

Compute

{"file name": "input.pdf", "number of pages": 29, "author": "Shiquan Zhang; Tianyi Zhang; Le Fang; Simon D'Alfonso; Hong Jia; Vassilis Kostakos", "title": "Do LLMs Need to See Everything? A Benchmark and Study of Failures in LLM-driven Smartphone Automation using Screentext vs. Screenshots"

Inference

{"file name": "input.pdf", "number of pages": 29, "author": "Shiquan Zhang; Tianyi Zhang; Le Fang; Simon D'Alfonso; Hong Jia; Vassilis Kostakos", "title": "Do LLMs Need to See Everything? A Benchmark and Study of Failures in LLM-driven Smartphone Automation using Screentext vs. Screenshots"

Hardware

{"file name": "input.pdf", "number of pages": 29, "author": "Shiquan Zhang; Tianyi Zhang; Le Fang; Simon D'Alfonso; Hong Jia; Vassilis Kostakos", "title": "Do LLMs Need to See Everything? A Benchmark and Study of Failures in LLM-driven Smartphone Automation using Screentext vs. Screenshots"

Evidence ids

Receipt path

/buildability/do-llms-need-to-see-everything-a-benchmark-and-study-of-failures-in-llm-driven-smartphone-automation-using-screentext-vs

Paper ref

do-llms-need-to-see-everything-a-benchmark-and-study-of-failures-in-llm-driven-smartphone-automation-using-screentext-vs

arXiv id

2604.17817

Freshness

Generated at

2026-04-21T02:39:59.163Z

Evidence freshness

stale

Last verification

2026-04-21T02:39:59.163Z

Sources

3

References

0

Coverage

50%

Hash state

Lineage hash

36efb30cf00a82e004ddf818f2a439f4a2573ff85349aff181132486014fa943

Canonical opportunity-kernel lineage hash.

Signature state

External signature

unsigned_external

No founder, registry, pilot, or production-adoption signature is attached to this receipt.

Verification

not_verified

Verification is blocked until an external signature is provided.

Blockers

  • Missing: repo_url
  • Missing: references
  • Missing: proof_status
  • Unknown: proof verification has not been recorded yet

Some score or evidence fields are outside the preferred freshness window.

repo_url

references

Missing proof, requirement, signature, approval, adoption, or telemetry fields are blockers and must not be inferred.

Truth Boundary

External gate remains unresolved for live deployment claims.

Buildability surfaces only report computed viability and proof receipts. They do not claim live production usage, pilot outcomes, founder sign-off, public Brier calibration, judge divergence, or external adoption unless explicitly sourced.