css: one-per-row Open-release cards on mobile (fix char-wrapped labels)

css: stack action buttons full-width on mobile (fix squished Run Research)

css: fix mobile-stack selector (Gradio 5.x has no gradio-container / .gr-column)

css: stack the two-column layout on mobile/tablet (fix broken phone layout)

search: parse knowledgeGraph/peopleAlsoAsk; treat 0-result as ok+hint

perf: enable Gradio queue concurrency (fix "long queue" serialization)

ui: header title -> "QUEST: Training Frontier Deep Research Agents with Fully Synthetic Tasks"

ui: Paper card default URL -> the final QUEST github-page site

agent: bump default MEMORY_TOKEN_THRESHOLD 16k -> 80k

ui: drop "Visited Sources" appendix from final answer

ui: hero heading verb back to "research" (deep research framing)

ui: fix hero heading grammar; default Max Turns 6 -> 15

diag: surface underlying httpx/ssl cause + endpoint/model in extractor_error

agent: Azure OpenAI support for visit extractor + condenser

agent: surface _llm_extract failure reason in visit tool_response

agent: wire research-repo Jina + LLM extractor + scholar + real condenser

ui: move connection-error tip from action-row to Output header

ui: auto-clear question after run; add connection-error retry tip

ui: capitalize QUEST in Space title, browser tab title, footer copy

ui: card-style resource links (paper/code/data/model) with icons; QUEST capitalization

ui: capitalize memory strategy labels; cap max_turns at 50

ui: rename Long-form example to Research Report; hyphenate memory labels

ui: capitalize QUEST in header and hero-heading

rename: Quest-35B -> QUEST-35B (uniform uppercase branding)

ui: label the condenser radio option as "(default)"

rename: Quest-4B -> Quest-35B (OSC vLLM serves the 35B checkpoint)

app: add QUEST_API_KEY for self-hosted endpoint Bearer

Settings: condenser first in radio + help; flatten Model wrapper so its label aligns with the rest

Bump base font to 17px and content width to 1700px for better legibility / less wasted side margin

Memory Strategy radio: drop selected-state highlight; only the dot signals selection

Memory Strategy: replace buggy Dropdown with Radio (4 stacked options); slider value text centered

Slider: flatten the number-input pill (no border) and hide the redundant reset button

Settings widgets: revert container=False, use elem_id with id-targeted CSS to flatten outer block while keeping labels

Use container=False on Settings widgets — official way to drop the outer block frame on Dropdown/Slider

Memory strategies: rename to upstream (condenser/discard_all/hide_tool_result), describe per inference code; quest-name letter-spacing matches paper microsite

Settings: equalize Textbox/Dropdown height, flatten Slider wrapper, default Temperature to 1.0

Quest brand mark: keep italic serif, drop terracotta color

Drop 'at Rice University' from Moon-speech example

Swap first example to Apollo 11 + JFK Moon speech (more iconic entities)

Header: drop Q mark, italicize Quest, capitalize byline; add vanilla memory strategy with explainer; restore example emojis

Hero: italic-serif Quest brand mark; harder BrowseComp-style examples

Shorten the first two examples to plainer well-known questions

UI polish: rename Space to Quest, harder examples, clean dropdown, hide progress bar

Raise endpoint request timeout to 600s (configurable via QUEST_REQUEST_TIMEOUT)

Force dropdown height to match the Model textbox

Match Memory Strategy dropdown height to the Model textbox

Reorder Settings: Model, Memory Strategy, Max Turns, Temperature

Streamline Settings panel: Quest-4B only, raise Max Turns, add Memory Strategy

Underline "built by OSU NLP Group" byline to cue the link

Update hero title and hide stray orange line above tab panel