Gemini 3.5 Flash

2026-10-16

RESEARCH · 1 SOURCE · InfoQ AI, ML & Data Engineering

Google open-sources EnvHarness to adapt training environments for self-evolving agents

Google researchers open-sourced EnvHarness, a framework that wraps static environments via the standard reset()/step() interface and applies composable setup, rule, and link mutations so environments adapt to an agent's weaknesses without changing the underlying environment or evaluators. The release also includes EnvRigger, an automated loop that analyzes agent rollouts to generate and validate targeted environment edits; Google reports consistent performance gains (≈2.9–3.7 percentage points) across benchmarks such as WebArena, SWE-bench Verified, and ALFWorld on models including Gemini, Qwen, and Claude Sonnet.

7.0

2026-10-15

RESEARCH · 1 SOURCE · arXiv cs.AI

Harness^2: two-level recursive framework for evolving LLM agent harnesses

Harness^2 is a plug-and-play recursive agent-harnessing framework that uses two linked recursion levels: task-level recursion (a contrastive Propose‑Probe‑Compose step) to edit a current task's harness without ground-truth verifiers, and harnessing-level recursion to extract experience and update a continual improvement harness for future tasks. Evaluated across three professional benchmarks and four model‑harness configurations, single-step edits improved GLM-5 and Gemini 3.5 Flash over their base harnesses, with reported gains up to 11.5 points on WorkBuddy-Bench, 4.5 points on JobBench, and 11.8 percentage points on LAB.

7.0

2026-10-04

RESEARCH · 1 SOURCE · The Decoder

Google researchers propose RRSI to stop agent harnesses memorizing tests

Google Cloud AI Research and academic collaborators introduce RRSI (Regularized Recursive Self-Improvement), a method that constrains automated harness rewrites so agentic systems avoid overfitting to benchmark tasks. Tested with a frozen Claude Opus 4.8 across eight benchmarks (including JobBench and ARC-AGI-3), RRSI improved training-task scores by up to 14.1 points, boosted performance on unseen tests by up to 4.7 points, and cut runtime token use by about 30%; the paper and code are on GitHub.

7.0

2026-10-02

RESEARCH · 1 SOURCE · arXiv cs.AI

Benchmarking automatic prompt optimization with a 1,118‑puzzle chess benchmark (arXiv:2610.00416v1)

The paper introduces a chess-based benchmark built from 1,118 Lichess puzzles to evaluate automatic prompt optimization (APO) for frozen LLMs, offering exact-match scoring, engine-based move evaluation, renewability, and low cost (the study ran for about $800). The authors evaluate six APO algorithms on eight target models (reporting that Gemini 3.5 Flash solves ~55% of puzzles), and release puzzles, code, and dataset-renewal scripts on GitHub.

6.0

2026-09-26

RESEARCH · 1 SOURCE · The Decoder

Study finds AI access nearly eliminates people saying “I don’t know”

Researchers running five experiments with 3,132 participants found that making a language model (Step 3.5 Flash) available dramatically reduced people’s willingness to withhold judgment—“I don’t know” responses fell from roughly 36–44% to about 3–6%—even though the model was often wrong on the tested movie-trivia items. Across studies, AI access raised confidence while cutting accuracy (correct answers fell from ~27.5% to ~9.2%); automatic display of AI answers produced similar effects, and modest financial incentives only partially mitigated the shift.

7.0

2026-06-24

MODELS · 1 SOURCE · Google DeepMind

Google DeepMind announces computer use in Gemini 3.5 Flash

Google DeepMind posted an item titled 'Introducing computer use in Gemini 3.5 Flash.' The supplied source text is not included, so the announcement appears to indicate a new 'computer use' capability for the Gemini 3.5 Flash model but no further details are available in the provided content.

6.0