FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W4311674792> ?p ?o ?g. }

• W4311674792 abstract "Learning from a limited number of experiences requires suitable inductive biases. To identify how inductive biases are implemented in and shaped by neural codes, we analyze sample-efficient learning of arbitrary stimulus-response maps from arbitrary neural codes with biologically-plausible readouts. We develop an analytical theory that predicts the generalization error of the readout as a function of the number of observed examples. Our theory illustrates in a mathematically precise way how the structure of population codes shapes inductive bias, and how a match between the code and the task is crucial for sample-efficient learning. It elucidates a bias to explain observed data with simple stimulus-response maps. Using recordings from the mouse primary visual cortex, we demonstrate the existence of an efficiency bias towards low-frequency orientation discrimination tasks for grating stimuli and low spatial frequency reconstruction tasks for natural images. We reproduce the discrimination bias in a simple model of primary visual cortex, and further show how invariances in the code to certain stimulus variations alter learning performance. We extend our methods to time-dependent neural codes and predict the sample efficiency of readouts from recurrent networks. We observe that many different codes can support the same inductive bias. By analyzing recordings from the mouse primary visual cortex, we demonstrate that biological codes have lower total activity than other codes with identical bias. Finally, we discuss implications of our theory in the context of recent developments in neuroscience and artificial intelligence. Overall, our study provides a concrete method for elucidating inductive biases of the brain and promotes sample-efficient learning as a general normative coding principle." @default.
• W4311674792 created "2022-12-28" @default.
• W4311674792 creator A5023195984 @default.
• W4311674792 creator A5039282308 @default.
• W4311674792 date "2022-12-16" @default.
• W4311674792 modified "2023-09-30" @default.
• W4311674792 title "Population codes enable learning from few examples by shaping inductive bias" @default.
• W4311674792 cites W1868105705 @default.
• W4311674792 cites W1961056875 @default.
• W4311674792 cites W1963769676 @default.
• W4311674792 cites W1966537851 @default.
• W4311674792 cites W1968942430 @default.
• W4311674792 cites W1977450749 @default.
• W4311674792 cites W1981025032 @default.
• W4311674792 cites W1986098213 @default.
• W4311674792 cites W1998050452 @default.
• W4311674792 cites W1998462024 @default.
• W4311674792 cites W2000089180 @default.
• W4311674792 cites W2009284521 @default.
• W4311674792 cites W2014158063 @default.
• W4311674792 cites W2017539895 @default.
• W4311674792 cites W2026928404 @default.
• W4311674792 cites W2029137641 @default.
• W4311674792 cites W2033705462 @default.
• W4311674792 cites W2036899314 @default.
• W4311674792 cites W2037843057 @default.
• W4311674792 cites W2038897596 @default.
• W4311674792 cites W2042925217 @default.
• W4311674792 cites W2043850729 @default.
• W4311674792 cites W2049977058 @default.
• W4311674792 cites W2051526952 @default.
• W4311674792 cites W2057736028 @default.
• W4311674792 cites W2063632045 @default.
• W4311674792 cites W2074423218 @default.
• W4311674792 cites W2075067380 @default.
• W4311674792 cites W2077381770 @default.
• W4311674792 cites W2079960018 @default.
• W4311674792 cites W2086841290 @default.
• W4311674792 cites W2093386234 @default.
• W4311674792 cites W2095087224 @default.
• W4311674792 cites W2100449221 @default.
• W4311674792 cites W2100483895 @default.
• W4311674792 cites W2103314569 @default.
• W4311674792 cites W2105430025 @default.
• W4311674792 cites W2105464873 @default.
• W4311674792 cites W2108992228 @default.
• W4311674792 cites W2108997568 @default.
• W4311674792 cites W2111306669 @default.
• W4311674792 cites W2119885245 @default.
• W4311674792 cites W2122066429 @default.
• W4311674792 cites W2122741244 @default.
• W4311674792 cites W2123713131 @default.
• W4311674792 cites W2124088405 @default.
• W4311674792 cites W2129375705 @default.
• W4311674792 cites W2140552585 @default.
• W4311674792 cites W2151035455 @default.
• W4311674792 cites W2156084351 @default.
• W4311674792 cites W2160654481 @default.
• W4311674792 cites W2160996222 @default.
• W4311674792 cites W2164365056 @default.
• W4311674792 cites W2166149038 @default.
• W4311674792 cites W2166206801 @default.
• W4311674792 cites W2167965189 @default.
• W4311674792 cites W2169191420 @default.
• W4311674792 cites W2170697082 @default.
• W4311674792 cites W2171281863 @default.
• W4311674792 cites W2286081412 @default.
• W4311674792 cites W2552737632 @default.
• W4311674792 cites W2588048611 @default.
• W4311674792 cites W2592738714 @default.
• W4311674792 cites W2621961142 @default.
• W4311674792 cites W2735683573 @default.
• W4311674792 cites W2766284800 @default.
• W4311674792 cites W2883327607 @default.
• W4311674792 cites W2885162916 @default.
• W4311674792 cites W2896846197 @default.
• W4311674792 cites W2923764619 @default.
• W4311674792 cites W2949895299 @default.
• W4311674792 cites W2952006560 @default.
• W4311674792 cites W2952921509 @default.
• W4311674792 cites W2954124637 @default.
• W4311674792 cites W2963305465 @default.
• W4311674792 cites W2963518130 @default.
• W4311674792 cites W2969343193 @default.
• W4311674792 cites W2975512858 @default.
• W4311674792 cites W2994737554 @default.
• W4311674792 cites W3004619146 @default.
• W4311674792 cites W3008714724 @default.
• W4311674792 cites W3011209639 @default.
• W4311674792 cites W3018252856 @default.
• W4311674792 cites W3134275903 @default.
• W4311674792 cites W3155379073 @default.
• W4311674792 cites W3159275413 @default.
• W4311674792 cites W3162003518 @default.
• W4311674792 cites W3171395320 @default.
• W4311674792 cites W3172995164 @default.
• W4311674792 cites W3201017316 @default.
• W4311674792 cites W3204632997 @default.
• W4311674792 cites W4210491714 @default.
• W4311674792 cites W4211049957 @default.

• next