###Participants
Participants were right-handed (assessed by Edinburgh Handedness Inventory) native born United States citizens (control for IAPS image self-report) who were medically healthy and exhibited no current Axis I psychopathology, including mood disorders, as assessed by the SCID-IV clinical interview. All participants reported no current use of psychotropic medication and produced a negative urine screen for drugs of abuse (cocaine, amphetamines, methamphetamines, marijuana, opiates, and benzodiazepines) immediately prior to the MRI scan. Participants also produced a negative urine screen prior to SCID-IV clinical interview. When necessary, participants’ vision was corrected to 20/20 using an MRI compatible lens system. Participants endorsing color blindness were excluded.

###Study Design
The study was performed over two separate sessions, each occurring on separate days. During Session 1, participants provided written informed consent, received screening for clinically relevant exclusionary criteria via structured clinical interview, and completed behavioral surveys and questionnaires. Neuroimaging and concurrent psychophysiological measurements were recorded during Session 2. Neuroimaging consisted of a T1-weighted scan, 2x Identify task fMRI scans, 1x Resting State task fMRI scan, and 2x Modulate task fMRI scans. The relevant task descriptions are elaborated below.

**System Identification Task (Identify)**
9.4 min task scan comprised of (n=90) affect processing induction trials and (n=30) affect self-regulation trials.  Affect processing induction was achieved by 2s presentation of image stimuli drawn from the International Affective Picture Set (IAPS), followed by an uniformly randomly sampled inter-trial stimulus of 2-6 s.  Affect self-regulation induction was achieved by cued-recall. On cued-recall trials, an IAPS image stimulus was presented for 2 s followed by a preparatory instruction (2 s of the word "FEEL" superimposed over the image), followed by a recall instruction, the word "FEEL" alone for 8s.  During recall the participant explicitly regulated their affect processing to match the affect processing induced by the cue image stimulus. Finally, a fixation cross replaced the word “FEEL” for an ITI sampled uniformly randomly from the range 2–6 s. Subjects were trained to regulate their affect upon seeing the recall instruction according to the following instructions. “When the image disappears and just the word ‘feel’ remains, we want you to re-imagine the image you just saw and try to re-feel how the image made you feel when you first saw it. Hold that feeling the entire time the word ‘feel’ is on the screen.” 

**Resting State Task (Rest)**
Subjects underwent 7.5 min of resting state scanning.  During training, subjects were instructed to “let your mind wander, not focusing on any specific thought” and to “try to keep your head still and your eyes open [and] … blink naturally.”  

**Intrinsic Neuromodulation of Core Affect Task (Modulate)**
10.3 min task consisting of (n=8) real-time guided affect self-regulation trials.  Each trial first presented an affective goal for 4s (positive valence, negative valence, high arousal, or low arousal).  The goal was followed by interleaved self-regulation instructions (the word "FEEL" for 10 s) and neurofeedback (4 s, a red dot on a grid representing the full range of valence and arousal scores possible).  During the self-regulation, subjects drew upon pre-prepared mental strategies (either imagined experiences or memories) in order to self-induce affect processing according to the desired goal. Real-time fMRI decoded valence and arousal processing (block-averaged over the preceding 10 s self-regulation period) were plotted on the grid as a red dot. The task used two randomly counterbalanced designs (A & B), which differed only in the order of the goals to be achieved. Each subject received both "real" and "sham" neurofeedback, randomly counterbalanced between subjects.  The "sham" neurofeedback was based on a high-performing subject during a pilot of the experiment design.

###Data Collection
We acquired all imaging data using the same Philips 3T Achieva X-series MRI scanner (Philips Healthcare, Eindhoven, The Netherlands) with a 32-channel head coil. We acquired anatomic images using an MPRAGE sequence (matrix = 256 x 256, 220 sagittal slices, TR/TE/FA = 8.0844/3.7010/8°, final resolution =0.94 x 0.94 x 1 mm3). We acquired functional images using the following EPI sequence parameters: TR/TE/FA = 2000 ms/30 ms/90°, FOV = 240 x 240 mm, matrix = 80 x 80, 37 oblique slices, ascending sequential slice acquisition, slice thickness = 2.5 mm with 0.5 mm gap, final resolution 3.0 x 3.0 x 3.0 mm3.

###Technical Note
The bold signal was acquired using the Philips Achieva seq+z slice timing with 37 slices. Based on TR=2.0 s, this sequence creates an inter-slice interval of 0.0541 s.  The following array has been copied into the SliceTime arrays of the template json files.

"SliceTiming": [0.0000,0.0541,0.1082,0.1623,0.2164,0.2705,0.3246,
	        0.3787,0.4328,0.4869,0.5410,0.5951,0.6492,0.7033,
		0.7574,0.8115,0.8656,0.9197,0.9738,1.0279,1.0820,
		1.1361,1.1902,1.2443,1.2984,1.3525,1.4066,1.4607,
		1.5148,1.5689,1.6230,1.6771,1.7312,1.7853,1.8394,
		1.8935,1.9476]